Abt Associates Inc.
4550 Montgomery Avenue
Bethesda, MD 20814
www. abtassociates. com
Modeled Attainment
Test Software
User's Manual
Prepared for
Office of Air Quality Planning and
Standards
U.S. Environmental Protection Agency
Research Triangle Park, NC
Brian Timin, Project Manager
June 2010
Prepared by
Abt Associates Inc.
-------
Table of Contents
Chapter 1 Welcome to MATS, the Modeled Attainment 9
Test Software
1.1 How to Use this Manual 10
1.2 Computer Requirements 11
1.3 Installing MATS 11
1.4 Installing an Updated Version of MATS, 14
1.5 Uninstalling MATS 14
1.6 Contact for Comments and Questions 15
16
Chapter 2 Terminology & File Types
2.1 Common Terms 16
2.1.1 ASR File 17
2.1.2 BMP File 17
2.1.3 Class I Area 17
2.1.4 Configuration File 17
2.1.5 CSV File 18
2.1.6 Deciviews 18
2.1.7 Design Value 19
2.1.8 Domain 19
2.1.9 Extinction 19
2.1.10 FRM Monitors 19
2.1.11 Gradient Adjustment 19
2.1.12 IMPROVE Monitors 19
2.1.13 Interpolation 20
2.1.14 Inverse Distance Weights 20
2.1.15 Log File 20
2.1.16 Output Navigator 21
2.1.17 Output File 21
2.1.18 Point Estimate 21
2.1.19 RRF 21
2.1.20 SANDWICH 21
2.1.21 Scenario Name 22
2.1.22 SMAT 22
2.1.23 Spatial Field 23
2.1.24 Spatial Gradient 23
2.1.25 STN Monitors 23
2.1.26 Temporal Adjustment 23
2.1.27 VNA 23
VNA - Detailed Description 24
2.2 File Types 27
Abt Associates Inc.
2
June 2010
-------
Contents
28
Chapter 3 Overview of MATS Components
3.1 Start 28
3.1.1 Annual PM Analysis 29
3.1.2 Daily PM Analysis 35
3.1.3 Ozone Analysis 42
3.1.4 Visibility Analysis 47
3.2 Output Navigator 50
3.3 Map View 52
3.4 Help 53
55
Chapter 4 Annual PM Analysis: Quick Start Tutorial
4.1 Step 1. Start MATS 55
4.2 Step 2. Output Choice 56
4.3 Step 3. Output Choice - Advanced 58
4.4 Step 4. Data Input 58
4.5 Step 5. Species Fractions Calculation Options 59
4.6 Step 6. Species Fractions Calculation Options-Advanced 61
4.7 Step 7. PM2.5 Calculation Options, 62
4.8 Step 8. Model Data Options 63
4.9 Step 9. Final Check 64
4.10 Step 10. Map Output 68
4.11 Step 11. View Output 76
80
Chapter 5 Annual PM Analysis: Details
5.1 Output Choice 81
5.1.1 Scenario Name 82
5.1.2 Standard Analysis 84
Step 1: Baseline Quartelry Average PM2.5 Calculation 84
Step 2: Baseline Quarterly Average Species Calculation 85
Step 3: Forecasted Quarterly Average Species Calculation 92
Step 4: Forecasted Design Value Calculation 94
Output Description 95
5.1.3 Quarterly Model Data 96
5.1.4 Species Fractions 98
Species Fractions Calculation 98
Output Description 99
5.2 Output Choice - Advanced 100
5.2.1 Spatial Field Estimates 102
Gradient-Adjustment- ("Fused fields") 103
Abt Associates Inc.
3
June 2010
-------
Output Description - Interpolate FRM & Speciation Monitor Data to Spatial
Field 105
Output Description - Interpolate Gradient-Adjusted FRM & Speciation
Monitor Data to Spatial Field 106
5.2.2 Miscellaneous Output 107
Quarterly Average Files 107
Output Description - High County Sites 111
Species Fractions Spatial Field 112
Output Description - Quarterly Average Speciated Monitors 114
Neighbor Files 114
5.3 Data Input 116
5.3.1 Species Data Input 118
5.3.2 PM2.5 Monitor Data Input 120
Unofficial Daily PM2.5 Monitor Data Input 120
Official Quarterly PM2.5 Monitor Data Input 122
5.3.3 Model Data Input 123
5.4 Species Fractions Calculation Options 124
5.4.1 Monitor Data Years 125
5.4.2 Delete Specifed Data Values 126
5.4.3 Minimum Data Requirements 127
5.5 Species Fractions Calculation Options - Advanced 129
5.5.1 Interpolation Options for Species Fractions Calculation 129
5.5.2 Miscellaneous Options 131
5.5.3 Internal Precision of the Calculations 131
5.6 PM2.5 Calculation Options 132
5.6.1 PM2.5 Monitor Data Years 133
5.6.2 Design Values 133
Completion Code Use 134
5.6.3 Valid FRM Monitors 135
5.6.4 NH4 Future Calculation 135
5.7 Model Data Options 135
5.8 Final Check 136
5.8.1 Running MATS in Batch Mode 138
139
Chapter 6 Daily PM Analysis: Quick Start Tutorial
6.1 Step 1. Start MATS 139
6.2 Step 2. Output Choice 140
6.3 Step 3. Output Choice - Advanced 142
6.4 Step 4. Data Input 142
6.5 Step 5. Species Fractions Calculation Options 143
6.6 Step 6. Species Fractions Calculation Options-Advanced 145
6.7 Step 7. PM2.5 Calculation Options, 146
6.8 Step 8. Model Data Options 147
6.9 Step 9. Final Check 148
6.10 Step 10. Map Output 152
6.11 Step 11. View Output 158
Abt Associates Inc.
4
June 2010
-------
Contents
161
Chapter 7 Daily PM Analysis: Details
7.1 Output Choice 161
7.1.1 Scenario Name 163
7.1.2 Standard Analysis 164
Step 1: Baseline Quarterly Peak PM2.5 Calculation 165
Step 2: Baseline Quarterly Peak Species Calculation 165
Step 3: Forecasted Quarterly Peak Species Calculation 173
Step 4: Forecasted Design Value Calculation 175
Output Description 177
7.1.3 Quarterly Peak Model Data 178
7.1.4 Species Fractions 180
Species Fractions Calculation 180
Output Description 180
7.2 Output Choice - Advanced 182
7.2.1 Miscellaneous Output 182
Quarterly Average Files 183
Output Description - High County Sites 185
Output Description - Quarterly Average Speciated Monitors 186
Neighbor Files 187
7.3 Data Input 188
7.3.1 Species Data Input 190
7.3.2 PM2.5 Monitor Data Input 192
Unofficial Daily PM2.5 Monitor Data Input 192
Official Daily PM2.5 Monitor Data Input 194
7.3.3 Model Data Input 195
7.4 Species Fractions Calculation Options 197
7.4.1 Monitor Data Years 198
7.4.2 Delete Specifed Data Values 199
7.4.3 Minimum Data Requirements 199
7.5 Species Fractions Calculation Options - Advanced, 201
7.5.1 Using Monitor Data to Calculate Species Fractions 202
7.5.2 Interpolation Options for Species Fractions Calculation 205
7.5.3 Miscellaneous Options 206
7.5.4 Internal Precision of the Calculations 207
7.6 PM2.5 Calculation Options 207
7.6.1 PM2.5 Monitor Data Years 208
7.6.2 Valid FRM Monitors 208
7.6.3 NH4 Future Calculation 209
7.7 Model Data Options 209
7.8 Final Check 210
7.8.1 Running MATS in Batch Mode 212
213
Chapter 8 Ozone Analysis: Quick Start Tutorial
8.1 Step 1. Start MATS 213
8.2 Step 2. Output Choice 214
Abt Associates Inc.
5
June 2010
-------
8.3 Step 3. Data Input 215
8.4 Step 4. Filtering and Interpolation, 216
8.5 Step 5. RRF & Spatial Gradient 218
8.6 Step 6. Final Check 219
8.7 Step 7. Load & Map Output 221
8.8 Step 8. View & Export Output 228
233
Chapter 9 Ozone Analysis: Details
9.1 Choose Desired Output 233
9.1.1 Scenario Name 234
9.1.2 Point Estimates 236
Baseline Ozone 236
Temporally-Adjust Baseline Ozone 237
9.1.3 Spatial Field 237
Baseline - interpolate monitor data to spatial field 238
Baseline - interpolate gradient-adjusted monitor data to spatial field 239
Forecast - interpolate monitor data to spatial field. Temporally-adjust ozone
levels 240
Forecast - interpolate gradient-adjusted monitor data to spatial field.
Temporally-adjust ozone levels 240
9.1.4 Ozone Output Variable Description 241
Ozone Monitors - monitor data, temporally adjusted 2015.csv 241
Ozone Monitors - county high monitoring sites, temporally adjusted
2015.csv 242
Spatial Field - interpolated monitor data, temporally adjusted;
gradient-adjusted monitor data, temporally adjusted 2015.csv 242
9.2 Data Input 243
9.2.1 Monitor Data 244
9.2.2 Model Data 245
EPA Default Model Data 246
9.2.3 Using Model Data 246
Nearby Monitor Calculation - Example 1 248
9.3 Filtering and Interpolation 249
9.3.1 Choose Ozone Design Values 250
9.3.2 Valid Ozone Monitors 251
Minimum Number Design Values 252
Max Distance from Domain 253
Required Design Values 254
9.3.3 Default Interpolation Method 255
9.4 RRF and Spatial Gradient 257
9.4.1 RRF Setup 258
RRF Calculation - Example 1 259
RRF Calculation - Example 2 261
RRF Calculation - Example 3 264
RRF Calculation - Example 4 266
RRF Calculation - Example 5 268
RRF Calculation Spatial Gradient with Backstop Threshold - Example 6 270
9.4.2 Spatial Gradient Setup 273
Spatial Gradient Calculation - Example 1 273
Abt Associates Inc.
6
June 2010
-------
Contents
Spatial Gradient Calculation - Example 2 275
Spatial Gradient Calculation - Example 3 277
9.5 Final Check 279
9.5.1 Running MATS in Batch Mode 280
281
Chapter 10 Visibility Analysis: Quick Start Tutorial
10.1 Step 1. Start MATS 281
10.2 Step 2. Output Choice 282
10.3 Step 3. Data Input 284
10.4 Step 4. Filtering 285
10.5 Step 5. Final_Check 286
10.6 Step 6. Load and Map Results 289
10.7 Step 7. Working with Configuration File 298
304
Chapter 11 Visibility Analysis: Details
11.1 Choose Desired Output 304
11.1.1 Scenario Name 305
11.1.2 Forecast Visibility at Class I Areas 307
Old IMPROVE Equation 309
New IMPROVE Equation 310
Choose Model Grid Cell 312
11.1.3 Visibility Output Variable Description 312
Forecasted Visibility Data.csv 313
Forecasted Visibility - all design values.csv 314
Class 1 Area and IMPROVE Monitor Identifiers and Locations.csv 315
Used Model Grid Cells - Base/Future Data.csv 316
11.2 Data Input 316
11.2.1 Monitor Data Input 317
Monitor Data Description (Old Equation) 318
Monitor Data Description (New Equation) 319
Linkage between Monitors & Class I Areas 321
11.2.2 Model Data Input 322
Using Model Data for Temporal Adjustment 324
11.3 Filtering 329
11.3.1 Example Valid Visibility Monitors 330
11.4 Final Check 332
11.4.1 Running MATS in Batch Mode 334
335
Chapter 12 Output Navigator
12.1 Add Output Files to Map 338
12.2 View Files 340
12.2.1 Configuration File 340
Abt Associates Inc.
7
June 2010
-------
12.2.2 Log File 342
12.2.3 Output Files 343
12.3 Extract Files, 347
350
Chapter 13 Map View
13.1 Loading Variables 350
13.1.1 Loading with Taskbar 352
13.2 Plotting a Value 355
13.2.1 Plotting Options 358
13.3 Zoom Options & Pan View 362
13.4 Standard Layers 364
13.5 Exporting Maps & Data Files 365
13.5.1 Exporting CSV Data File 367
369
Chapter 14 Frequently Asked Questions
14.1 Error: MATS will not create a folder for extracting files 369
14.2 Where is there a description of output variables? 369
14.3 Removing Data 369
371
Chapter 15 References
Abt Associates Inc.
8
June 2010
-------
Welcome to MATS, the Modeled Attainment Test Software
1 Welcome to MATS, the Modeled Attainment
Test Software
The Modeled Attainment Test Software (MATS) is primarily intended as a tool to
implement the modeled attainment tests for particulate matter (PM2 5) and ozone (03), and
to perform the uniform rate of progress analysis for regional haze (visibility). Detailed
information on the attainment tests can be found in U.S. EPA's modeling guidance,
"Guidance on the Use of Models and Other Analyses for Demonstrating Attainment of the
Air Quality Goals for Ozone, PM2 5, and Regional Haze." The modeling guidance can be
found at http://www.epa.gov/ttn/scram/guidance sip.htm.
This Chapter provides a brief description of how to use this manual, computer
requirements, steps to install and uninstall MATS, and contact information for comments
and questions:
How to Use this Manual
Computer Requirements
Installing MATS
Uninstalling MATS
Contact for Comments and Questions.
Abt Associates Inc.
9
June 2010
-------
Welcome to MATS, the Modeled Attainment Test Software
Umats
Help T
Start Map View Output Navigator
Ozone Analysis
Visibility Analysis
Annual PM Analysis
Daily PM Analysis
1.1 How to Use this Manual
This manual provides step-by-step instructions on how to use MATS.
New users should start with the Overview of MATS Components chapter, which is very
short, but provides a good overview of the model and how it works. You can then use
tutorial chapters to get started using the model. There are separate tutorials for Annual
and Daily Particulate Matter (PM), Ozone, and Visibility. In addition to these relatively
simple tutorials, you can go on to learn more on each subject in the chapters on Annual PM
Analysis: Details. Daily PM Analysis: Details. Ozone Analysis: Details, and Visibility
Analysis: Details. Use the rest of the manual to answer any specific questions you may
have. There is a chapter on the Output Navigator, which is the starting point for examining
your results. The Map View chapter details how to map results. Finally, the Frequently
Asked Questions chapter reviews and answers some of the common questions that arise
when using MATS.
In sections that provide instructions on navigating the model, the following conventions
are observed: menu items, buttons, and tab and selection box labels are in bold type;
prompts and messages are enclosed in quotation marks; and drop-down menu items,
options to click or check, and items that need to be filled in or selected by the user are
italicized. Common terms are defined in the Terminology and File Types chapter. The
Reference section provides citations for documents relevant to MATS.
Abt Associates Inc.
10
June 2010
-------
Welcome to MATS, the Modeled Attainment Test Software
1.2 Computer Requirements
MATS requires a Windows platform, and can be used on machines running Windows2000,
as well as more recent versions of Windows. In particular, MATS requires a computer
with:
Windows 2000 or greater.
512 megabytes of RAM or greater.
Intel® or compatible processor, Pentium 166 MHz or higher. 1 GHz processor or
greater recommended for optimum performance.
A CD-ROM drive for CD based installation. Alternatively, a high speed internet
connection can be used to download the installer. The installer package can be found at:
At least 3 GB free space recommended.
1.3 Installing MATS
Load the installation file (MATS_Setup.exe) onto your hard drive. Double-click the file.
This will initiate the installation process, which takes about five to ten minutes, depending
on the speed of your computer.
Abt Associates Inc.
11
June 2010
-------
Welcome to MATS, the Modeled Attainment Test Software
#MATS - InstallShield Wizard
X
< Back-
Next >
Cancel
Click the Next button. This will bring up the MATS - InstallShield Wizard.
Abt Associates Inc.
12
June 2010
-------
Welcome to MATS, the Modeled Attainment Test Software
vS'MATS - InstallShield Wizard
Ready to Install the Program
The wizard is ready to begin installation.
If you want to review or change any of your installation settings, click Back. Click Cancel to exit
the wizard.
Current Settings:
Setup Type:
Typical
Destination Folder:
C:\Program Files\Abt Assoc iates\JvlATS\
User Information:
Name: Don
Company: Don
InstallShield
Click the Install button. After the installation of MATS, a final window will appear to
complete the process.
Abt Associates Inc.
13
June 2010
-------
Welcome to MATS, the Modeled Attainment Test Software
#MATS - InstallShield Wizard
X
InstallShield Wizard Completed
The InstallShield Wizard has successfully installed MATS. Click
Finish to exit the wizard,
[v] Launch the program
Finish
Cancel
Click the Finish button.
Note that some problems have occurred in the past, when trying to install MATS from a
network drive. If this problem occurs, move the MATS_Setup.exe file to your local hard
drive.
1.4 Installing an Updated Version of MATS
If a previous version of MATS is already installed on your computer, you will need to
uninstall the old version using the Windows Control Panel prior to installing the new
version (see next section). Note that uninstalling MATS will not delete your MATS output
files.
1.5 Uninstalling MATS
To uninstall MATS, go to Control Panel, Add/Remove Programs and highlight MATS.
Abt Associates Inc.
14
June 2010
-------
Welcome to MATS, the Modeled Attainment Test Software
&Add or Remove Programs
Gg
•3£p
Change or
Remove
Programs
ft
Add New
Programs
&
Set Program
Access and
Defaults
Add/Remove
Windows
Components
Currently installed programs:
~ Show updates Sort by: Name
i0 iTunes
Size
47.74MB A
is' Java 2 Runtime Environment, SE vl.4.2_03
Size
107.00MB
H MATS
Size 1.622.00MB 1
Click here for suoDort information.
Used
rarelv 11
To change this program or remove it from your computer, click Change or
Remove.
ibbP
d McAfee SecurityCenter
["] McAfee VirusScan
Microsoft ,NET Framework 1.1
Microsoft .NET Framework 1,1 Hotfix (KB886903)
Microsoft Compression Client Pack 1,0 for Windows XP
31 Microsoft Office XP Standard
Size
477.00MB
# Microsoft User-Mode Driver Framework Feature Pack 1.0
MSN Music Assistant
# MSXML 4.0 SP2 (KB927978)
Size
2.56MB
03 DnDf-R in
Ci-7^
T nc nnn/ip v
Click the Remove button. This will bring up a window asking you to confirm the removal.
Add or Remove Programs
Note that removing the software will not remove the files that you have generated with
MATS. For example, the Output folder will remain with any files (e.g., *.ASR files) that
you have created.
1.6 Contact for Comments and Questions
For comments and questions, please contact Brian Timin at the United States
Environmental Protection Agency.
Address: C339-01, USEPA Mailroom, Research Triangle Park, NC 27711
Email: timin.brian@epa.gov
Telephone: 919-541-1850.
Abt Associates Inc.
15
June 2010
-------
Terminology & File Types
2 Terminology & File Types
The first section of this chapter explains Common Terms used in this user's manual and in
the model, and references, where possible, other sections in this manual to find more
detailed information. The second section describes the File Types used in MATS.
2.1 Common Terms
The following include terms commonly used in MATS:
ASR File
BMP File
Class I Area
Configuration File
CSV File
Deci views
Design Value
Domain
FRM Monitors
Gradient Adjustment
IMPROVE Monitors
Interpolation
Inverse Distance Weights
Log File
Output Navigator
Output File
Point Estimate
RRF
SANDWICH
Scenario Name
SMAT
Spatial Field
Abt Associates Inc.
16
June 2010
-------
Terminology & File Types
Spatial Gradient
STN Monitors
Temporal Adjustment
VNA
2.1.1 ASR File
An ASR File contains three types of results from a MATS run: Log File. Configuration
File, and Output Files. The extension .ASR is used after the Scenario Name. The data in
an .ASR file can viewed and extracted using the Output Navigator.
2.1.2 BMP File
BMP is a standard file format for computers running the Windows operating system. The
format was developed by Microsoft for sotoring bitmap files in a device-independent
bitmap (DIB) format that will allow Windows to display the bitmap on any type of display
device. The term "device independent" means that the bitmap specifies pixel color in a
form independent of the method used by a display to represent color.*
* See: http://www.prepressure.com/formats/bmp/fileformat.htm.
2.1.3 Class I Area
A Class I Area is defined by the Clean Air Act to include national parks greater than 6,000
acres, wilderness areas and national memorial parks greater than 5,000 acres, and
international parks that existed as of August 1977.* The Regional Haze rule requires
visibility improvements in 156 specific Class I areas. The MATS visibility analysis will
calculate visibility values for these areas.
* See: http://vista.cira.colostate.edu/views/Web/General/Glossarv.aspx.
2.1.4 Configuration File
A Configuration File stores the choices that you have made when using MATS. A useful
feature of a Configuration File is that it is reusable. You can use an existing Configuration
File, make some minor changes to generate a new set of results, without having to
explicitly set each of the choices you made in the previous Configuration. The section on
the Output Navigator provides additional details on accessing and viewing a Configuration
Abt Associates Inc.
17
June 2010
-------
Terminology & File Types
File.
2.1.5 CSV File
Is a comma separated values (CSV) file (*.csv) which can be read using a text editor, or by
various spreadsheet and database programs, such as Microsoft Excel.
Note: Detailed formatting in .csv files such as leading zeroes and "" cannot be seen in
Excel. To see formatting of MATS input files, open .csv files with a text editor, such as
WordPad.
9 OZONE_ASIP_input.csv - WordPad
. nx
File Edit View
Insert Format Help
~ m
Ek #4 Jt % 4 «o 1%
|DesignValue
/V
ID, TYPE,
LAT, LONG, POC, DVYEAR, 03, STATE NAME,
COUNTY NAME
"010030010"
,30 . 497778,-87.881389,1,2001,-9,
"Alabama"
"Baldwin"
"010030010"
,30 . 497778,-87.881389,1,2002, 82,
"Alabama"
"Baldwin"
"010030010"
,30 . 497778,-87.881389,1,2003, 76,
"Alabama"
"Baldwin"
"010030010"
,30 . 497778,-87.881389,1,2004, 76,
"Alabama"
"Baldwin"
"010270001"
,33 .281111,-85.802222,1,2001, 84,
"Alabama"
"Clay"
"010270001"
,33.2 81111,-85.802222,1,2 0 02,82,
"Alabama"
"Clay"
"010270001"
,33.281111,-85.802222,1,2003,80,
"Alabama"
"Clay"
"010270001"
,33.281111,-85.802222,1,2004,76,
"Alabama"
"Clay"
"010510001"
,32 .498333,-86.136667,1,2001, 79,
"Alabama"
"Elmore"
"010510001"
,32 .498333,-86.136667,1,2002, 80,
"Alabama"
"Elmore"
"010510001"
,32 .498333,-86.136667,1,2003, 76,
"Alabama"
"Elmore"
"010510001"
,32 .498333,-86.136667,1,2004, 74,
"Alabama"
"Elmore"
"010550011"
,33 .9039,-86.0539,1,2001,-9,"Alabama","Etowah"
"010550011"
,33.9039,-86.0539,1,2002,-9,"Alabama","Etowah"
"010550011"
,33 .9039,-86.0539,1,2003,-9,"Alabama","Etowah"
"010550011"
,33.9039,-86.0539,1,2004,75,"Alabama","Etowah"
"010730023"
,33 .553056,-86.815,1,2001,-9,"Alabama","Jefferson"
"M n73nn?3".
.33 553ri5f=; .-R6 fi 1 5 . 1 . ? fl fl? . R? . "a 1 ahama " . ".Tpf f Prmn "
V
<
>
For Help, press F1
NUM
2.1.6 Deciviews
A third measure of visibility is the deciview index, which EPA selected as the standard
metric for tracking progress in EPA's regional haze program, largely because it provides a
linear scale for perceived visual changes over a wide range of conditions. 1 On a particle-
free, pristine day, the deciview index has a value of zero (SVR=391 km). On a relatively
clear day in the Great Smoky Mountains the deciview index might be about 16 (SVR=79
km) and on a relatively hazy day the deciview index might be about 31 (SVR=201 km).
For each 10 percent increase in light-extinction, the deciview index goes up by one. So,
higher deciview values mean worse visibility. Under many scenic conditions, a change of
one deciview is considered to be just perceptible by the average person.
Abt Associates Inc.
18
June 2010
-------
Terminology & File Types
2.1.7 Design Value
The monitored reading used by EPA to determine an area's air quality status; e.g., for
ozone, the 3 year average of the annual fourth highest reading measured at each monitor is
the design value. Ozone design values are calculated in accordance with 40 CFR Part
50.10, and Appendix I to Part 50. The calculation of annual and 24-hour average PM2 5
design values can be found in 40 CFR Part 50, Appendix N.
2.1.8 Domain
A Domain (or Model Domain) refers to the coverage of an air quality model, or the area of
the country for which there are model values. MATS calculates design values and/or
spatial fields for an area encompassed by the coordinates given within a MATS input file.
2.1.9 Extinction
Light extinction is the sum of the light scattering and light absorption by particles and
gases in the atmosphere, and is measured in inverse megameters (Mm-1), relating how
much light is extinguished per megameter. Higher extinction values mean worse visibility.
2.1.10 FRM Monitors
Federal Reference Method (FRM) monitors used to determine attainment or
nonattainment.The term "FRM" is frequently used to describe the network of PM2 5 mass
monitors.
2.1.11 Gradient Adjustment
A gradient adjustment is used to scale, or adjust, monitor data when using monitor data to
estimate air pollution levels in unmonitored areas. It is calculated as the ratio of the model
value in the unmonitored area to the value in the monitored area. In MATS, gradient
adjustments can be used for PM Analyses and Ozone Analyses.
2.1.12 IMPROVE Monitors
Interagency Monitoring of PROtected Visual Environments (IMPROVE) is a collaborative
monitoring program established in the mid-1980s. IMPROVE objectives are to provide
data needed to assess the impacts of new emission sources, identify existing man-made
visibility impairment, and assess progress toward the national visibility goals that define
protection of the 156 Class I areas.*
* See: http://vista.cira.colostate.edu/views/Web/General/Glossarv.aspx.
Abt Associates Inc.
19
June 2010
-------
Terminology & File Types
2.1.13 Interpolation
Interpolation is the process of estimating the air quality level in an unmonitored area by
using one or more nearby air quality monitors. The technique used in MATS is called
Voronoi Neighbor Averaging CVNAY
2.1.14 Inverse Distance Weights
The weight given to any particular monitor is inversely proportional to its distance from
the point of interest.
Example, Inverse Distance Weights
Assume there are four monitors (A, B, C, and D) that are a varying distance from a point E.
Assume the distances are 10, 15, 15, and 20 kilometers respectively. The weights will be
as follows:
WeightA= 10/(10+15+15+20)= 10/60 = 0.17
WeightB = 15 / (10+15+15+20) = 0.25
Weightc = 15 / (10+15+15+20) = 0.25
WeightD = 20 / (10+15+15+20) = 0.33
Example, Inverse Distance Squared Weights
Assume there are four monitors (A, B, C, and D) that are a varying distance from a point E.
Assume the distances are 10, 15, 15, and 20 kilometers respectively. The weights will be
as follows:
WeightA= 10/(102+152+152+202)= 100/950 = 0.11
WeightB = 15 / (102+152+152+202) = 0.24
Weightc = 15 / (102+152+152+202) = 0.24
WeightD = 20 / (102+l52+l52+202) = 0.42
2.1.15 Log File
A Log File provides information on a variety of technical aspects regarding how a results
file (*.ASR) was created. This includes the version of MATS, the date and time the *.ASR
file was created.
Abt Associates Inc.
20
June 2010
-------
Terminology & File Types
Start Map View Output Navigator | Run Log
Close
»» Start MATS.exe v 1.1.0.-1 2007-02-25 22:14:00
Starting iteration 0
Loading Default membership file...0.086 s.
Loading wind profiles file...0.026 s.
Loading Ozone monitor data...0.228 s.
WARNING: Base year of modeling changed to agree with Ozone data
Loading Baseline Model Data...87.966 s.
Calculating metric for Gradients...7.610 s.
Interpolating to spatial fields...13.510 s.
Reading future modeling file: C:\Program Files\AbtAssociates\MATS\SampleData\ozone_model_data_2015.csv...91.075 s.
Running future year estimates at monitors...1.040 s.
Spatial interpolations to model cells...53.145 s.
Total execution time: 262.823 s.
«« Stop MATS.exe 2007-02-25 22:18:24
2.1.16 Output Navigator
The Output Navigator allows you to load results files that you have previously created.
You can then view these data in maps and in tables, or export the data to text files, which
you can then load into a program such as Excel. Additional details are in the Output
Navigator Chapter.
2.1.17 Output File
An Output File is one of the file types within a *.ASR results file. The types of Output
Files available depend on the type of analysis (PM, Ozone, or Visibility) and the output
choices that you have specified in the Configuration File.
2.1.18 Point Estimate
A calculation within MATS that is performed at (or near) the location of ambient air
monitors. The output files will contain base and/or future year results at each valid
monitoring location.
2.1.19 RRF
The relative response factor is the ratio of the future year modeled concentration predicted
near a monitor (averaged over multiple days) to the base year modeled concentration
predicted near the monitor (averaged over the same days).
2.1.20 SANDWICH
The SANDWICH process is used to adjust STN and IMPROVE monitor data so that it is
Abt Associates Inc.
21
June 2010
-------
Terminology & File Types
consistent with FRM monitor data. SANDWICH stands for Sulfates, Adjusted Nitrates, D
erived Water. Inferred Carbonaceous mass, and estimated aerosol acidity (H+).*
* For more details, see: Frank, N., 2006: "Retained Nitrate, Hydrated Sulfates, and Carbonaceous
Mass in Federal Reference Method Fine Particulate Matter for Six Eastern U.S. Cities" J. Air Waste
Mange. Assoc., 56, 500-511.
2.1.21 Scenario Name
The name given to a set of results generated by MATS. The Scenario Name is used in
several ways: (1) the results file (*.ASR) uses the Scenario Name; (2) an output folder,
containing results extracted from a *.ASR file, is given the Scenario Name; and (3) the
Output File names begin with the Scenario Name.
The Scenario Name is specified when choosing the desired output, such as in the case of an
ozone analysis.
2.1.22 SMAT
The Speciated Modeled Attainment Test (SMAT) is used to forecast PM2 5 values. The
main steps are as follows:
Abt Associates Inc.
22
June 2010
-------
Terminology & File Types
Derive quarterly mean concentrations for each component of PM2 5 by multiplying FRM
PM2 5 by fractional composition of each species;
Calculate a model-derived relative response factor fRRF) for each species;
Multiply each RRF times each ambient PM2 5 component (for each quarter) to get the
future concentrations;
Sum the future quarterly average components; and
Average the four mean quarterly future PM2 5 concentrations.
2.1.23 Spatial Field
A Spatial Field refers to air pollution estimates made at the center of each grid cell in a
specified modeling domain. For example, MATS might calculate ozone design values for
each grid cell in the modeling domain. Several types of Spatial Fields can be calculated
for ozone and PM. (See the sections for ozone and PM for additional details.)
2.1.24 Spatial Gradient
A Spatial Gradient is the ratio of mean model values at an unmonitored location over the
mean model values at a monitor. Spatial Gradients can be used in the calculation of
Spatial Fields for ozone and PM. (See the sections for ozone and PM for additional
details.)
2.1.25 STN Monitors
In meeting the requirements to monitor and gather data on the chemical makeup of fine
particles, EPA established a Speciation Trends Network (STN). These STN monitors were
placed at various national air monitoring stations (NAMS) and State and local air
monitoring stations (SLAMS) across the Nation.
2.1.26 Temporal Adjustment
A temporal adjustment refers to multiplying ambient monitor data with a model derived
relative response factor fRRF) in order to generate an estimated future year concentration.
2.1.27 VNA
Voronoi Neighbor Averaging (VNA) is an algorithm used by MATS to interpolate air
quality monitoring data to an unmonitored location. MATS first identifies the set of
monitors that best "surround" the center of the population grid cell, and then takes an
inverse-distance weighted average of the monitoring values.
Abt Associates Inc.
23
June 2010
-------
Terminology & File Types
2.1.27.1 VNA - Detailed Description
Voronoi Neighbor Averaging (VNA) algorithm uses monitor data directly or in
combination with modeling data. MATS first identifies the set of monitors that best
"surround" the point of interest.
*
*
*
#
*
*
*
*
# = Center Grid-Cell "E"
•k
= Air Pollution Monitor
In particular, MATS identifies the nearest monitors, or "neighbors," by drawing a polygon,
or "Voronoi" cell, around the center of the point of interest. The polygons have the special
property that the boundaries are the same distance from the two closest points.
Abt Associates Inc.
24
June 2010
-------
Terminology & File Types
*
= Air Pollution Monitor
MATS chooses those monitors that share a boundary with the center of grid-cell "E."
These are the nearest neighbors, we use these monitors to estimate the air pollution level
for this grid-cell.
Abt Associates Inc.
25
June 2010
-------
Terminology & File Types
*
= Air Pollution Monitor
To estimate the air pollution level in each grid-cell, MATS calculates an inverse-distance
weighted average of the monitor values. The further the monitor is from the grid cell, the
smaller the weight. In the figure below, the weight for the monitor 10 miles from the
center of grid-cell E is calculated as follows:
1
d — -0 27
a+±+±i
^20 16 14/
The weights for the other monitors are calculated in a similar fashion. MATS then
calculates an inverse-distance weighted average for grid-cell E as follows:
Estimate = 0.35*80 ppb + 0.24*90 ppb+ 0.24*60 ppb + 0.18*100 ppb = 81.2 ppb
Abt Associates Inc.
26
June 2010
-------
Terminology & File Types
A
B
Monitor:
1995 90 ppb ^
15 miles
C
*
D
Monitor: *
1995 80 ppb
10 miles
HI
F
*
Monitor:
1995 60 ppb
15 miles
G
*
/ H
*
Monitor:
1995 100 ppb
20 miles
I
*
# = Center Grid-Cell "E"
= Air Pollution Monitor
2.2 File Types
The primary results file generated by MATS has a ASR extension, which is specific to
MATS. To view the results you have generated in other programs (e.g., MS Excel), you
can export .CSV files using the Output Navigator.
Abt Associates Inc.
27
June 2010
-------
Overview of MATS Components
3 Overview of MATS Components
Upon starting MATS for the first time, you will see the following main window.
Umats
Help
Start
M ap Vi e w Outp ut Navi g ato r
Ozone Analysis
Visibility Analysis
Annual PM Analysis
Daily PM Analysis
Stop Info
There are three main tabs: Start, Map View, and Output Navigator. The Start tab allows
you to calculate Annual and Daily PM, Ozone and Visibility levels. The Map View tab
allows to map your results. The Output Navigator tab allows you to view your results
either as tables or maps. Finally, the Help menu at the top of the main window provides
explanations and examples of all of the functionality in MATS.
This Chapter gives a brief description of each of these items. All of these topics are
covered in greater detail in subsequent chapters of this manual.
3.1 Start
The Start tab gives you the choice to analyze Particulate Matter (PM). Ozone or Visibility.
To begin, click on one of the three buttons.
One of the key features of MATS is the Configuration. This is a reusable file that stores
the choices that you have made when using MATS. You can use an existing Configuration
File, make some minor changes to generate a new set of results, without having to
explicitly set each of the choices you made in the previous Configuration.
Abt Associates Inc.
28
June 2010
-------
Overview of MATS Components
When you click on one of the analysis buttons, you will be asked whether you want to
create a new Configuration, or whether you want to use an existing Configuration.
Configuration Management
(* plreate New Configuration!
C Open Existing Configuration
Cancel
Go
Make your choice and then click Go. MATS will then take you through a series of
windows specifying the options available for each analysis.
3.1.1 Annual PM Analysis
With the Standard Analysis, MATS can forecast annual PM2.5 design values at monitor
locations. MATS can also calculate quarterly model data files and a species fractions file.
The Choose Desired Output window lets you specify the type of calculation(s) that you
would like MATS to perform. These different assumptions are discussed in the Output
Choice section of the Annual PM Analysis: Details chapter.
Choose Desired Output
Scenario Name: Example Annual F'M
Standard Analysis
p Interpolate monitor data to FRM monitor sites. Temporally-adjust.
Quarterly Model Data
W Output quarterly average model data file,
p Output used quarterly average model data file.
Species Fraction
[Output species fractions filej
Actions on run completion
Automatically extract all selected output files
Annual PM Analysis
Choose Desired Output
¦ Output Choices - Advanced
¦ Data Input
¦ Species Fractions Options
¦ S pecies Fractions - Advanced
¦ PM2.5 Calculation Options
¦ Model Data Options
Final Output and Check
< Back
Next >
In the Output Choice Advanced window, MATS lets you choose from among two main
options: Spatial Field Estimates and Miscellaneous Output that is generally used for
quality assurance (QA). Within each of these two main options there are a number of
Abt Associates Inc.
29
June 2010
-------
Overview of MATS Components
choices. Details regarding these choices are in the Output Choice - Advanced section of
the Annual PM Analysis: Details chapter.
Annual PM Analysis
I Choose Desired Output
Output Choices - Advanced
Species Fractions Options
Species Fractions - Advanced
PM2.5 Calculation Options
Model Data Options
Final Output and Check
Output Choice - Advanced
Spatial Field Estimates
Forecast
|7 Interpolate FRM and speciation monitor data to spatial field. Temporally adjust.
|7 Interpolate gradient-adjusted FRM and speciation monitor data to spatial field. Temporally adjust.
Miscellaneous Outputs
Quarterly average files
|7 Point 17 Spatial Field
|7 Spatial Field - gradient-adjusted
High county sites
|7 File "C"
Species fractions spatial field
|7 Spatial Field |7 Spatial Field - gradient-adjusted
Quarterly average speciated monitors
|7 File "E"
Neighborfiles
|7 Point
[7 iSpatial Field
< Back
Next >
Cancel
In the Data Input window, you specify the MATS input files that are used in each scenario.
There are three main types of files which must be specified. These include ambient
PM2.5 species data, ambient total PM2.5 data (FRM and IMPROVE), and gridded model
output data (e.g. CMAQ or CAMx data). There is specific terminology that is used on the
Data Input page. "Official" data refers to PM2.5 FRM data that can be used to determine
official design values for compliance purposes (comparison to the NAAQS). Other
datasets which may not have rigid regulatory significance are sometimes referred to as
"unofficial" data. The format for the data is in the Data Input section of the Annual PM
Analysis: Details chapter.
Abt Associates Inc.
30
June 2010
-------
Overview of MATS Components
Annual PM Analysis
]
¦ Choose Desired Output
¦ Output Choices-Advanced
Data Input
¦ Species Fractions Options
¦ Species Fractions - Advanced
¦ PM 2.5 Calculation 0 ptions
¦ Model Data Options
¦ Final Output and Check
Data Input
Species Data
® IS cedes MonitorD ata File! | C: \Program FilesSAbt Associates\M AT S \S ampleD ata\S pecies-f
O Species Fractions File | point
| spatial field ^
PM2.5 Monitor Data
Unofficial Dailv Averaae PM2.5 Data File ffor All Species Fractions & PM2.5 Spatial Field!
| C: \Program FilesSAbt AssociatesSM AT S \S ampleD ataSPM 25-f or-f ractions-020G-v2. csv
Official Quarterly Average FRM Data File (for PM2.5 Point Calculations)
| C: \Program FilesSAbt AssociatesSM AT S SS ampleD ataSAnnual-official-FR M -99-07-v2. csv
Model Data
® Daily model data input O Quarterly model data input
B aseline File | C: SProgram FilesSAbt AssociatesSM AT S SS ampleD ataS2002cc_E U S_PM 25_sub. csv
Forecast File |C:\Program FilesSAbt AssociatesSMATSSSampleDataS2020cc_EUS_PM25_sub.csv
< Back Next > Cancel
The Species Fractions Calculation Options has two main sections. One involving
speciated monitor data (e.g., STN and IMPROVE monitors) and the other total PM2.5
monitor data (FRM and IMPROVE). For each type of data you can specify the years of
interest, whether you want to delete certain data, and the minimum amount of data for a
monitor to be considered "valid" (and thus included in the calculations). Details on these
options are in theSpecies Fractions Calculation Options section of the Annual PM
Analysis: Details chapter.
Abt Associates Inc.
31
June 2010
-------
Overview of MATS Components
Annual PM Analysis
BHSm Choose Desired Output
¦ Output Choices - Advanced
¦ Data Input
Species Fractions Options
¦ Species Fractions - Advanced
¦ PM2.5 Calculation Options
¦ Model Data Options
Final Output and Check
Species Fractions Calculation Options
IMPROVE-STN Monitor Data
Monitor Data Years
Start Year End Year
12002
_^J [2004 ~T\
Delete Specified Data Values
W EPA-specified deletions from monitor data
User-specified deletions from monitor data
Minimum Data Requirements
Minimum number of valid days per quarter
Minimum number of valid years required for valid season
Minimum number of valid seasons for valid monitor
PM2.5 Monitor Data
Monitor Data Years
Start Year End Year
[2002 3 3
Delete Specified Data Values
W EPA-specified deletions from monitor data
User-specified deletions from monitor data
Minimum Data Requirements
Minimum number of valid days per quarter
Minimum number of valid years required for valid season
Minimum number of valid seasons for valid monitor (point calculations)
Minimum number of valid seasons for valid monitor (spatial fields calculations) |~~
Trg
r
< Back
Next >
Cancel
The Species Fractions Calculation Options - Advanced screen allows you to make
relatively advanced choices for your analysis. Generally speaking, the default options
settings are consistent with the EPA modeling guidance document (note: the start and end
years should always be set to match the relevant base modeling year). One set of options
allows you to specify the interpolation weighting that you want to use and whether the
interpolation involves a maximum distance or not. The second set of options involves
choices regarding ammonium, blank mass, and organic carbon. Details on these options
are in the Species Fractions Calculation Options - Advanced section of the Annual PM
Analysis: Details chapter.
Abt Associates Inc.
32
June 2010
-------
Overview of MATS Components
Annual PM Analysis
Choose Desired Output
¦ Output Choices - Advanced
¦ Data Input
¦ Species Fractions Options
Species Fractions - Advanced
¦ PM 2.5 Calculation 0 ptions
¦ Model Data Options
Final Output and Check.
Species Fractions Calculation Options - Advanced
Interpolation Options
PM2.5 |~ ^ Crustal [inverse Distance Squared V | 9IJ0LI0 *>
S04 | Inverse Distance Squared V ~ [ [SOOOCT^ DON Inverse Distance Squared V ~ | 190000 SI
N03 | Inverse Distance Squared V - 00000 OC | Inverse Distance Squared V 190000 ** :0
EC | Inverse Distance Squared V ~] 190000 ** NH 4 | n1verse Distance Squared V ~
Salt | Inverse Distance Squared V ~] |90000 «\~v] »|
Miscellaneous Options
AmmGnium
® Use DON values
O Use measured ammonium
NH4 percentage evaporating (0-100) I 53
Default Blank Mass
Default Blank Mass
Organic Carbon
Organic carbon mass balance floor
Organic carbon mass balance ceiling I 5s=j
< Back
Next >
The PM2.5 Calculation Options window allows you to specify the particular years of
monitor data that you want to use from the input file you specified in the Data Input
section. You can specify whether to use "official" or "custom" design values and whether
monitors should have a minimum number of design values or a design value for a
particular year. You can also specify how to calculate future NH4 levels. Details on these
options are in the PM2.5 Calculation Options section of the Annual PM Analysis: Details
chapter.
Abt Associates Inc.
33
June 2010
-------
Overview of MATS Components
Annual PM Analysis
PM2.5 Calculation Options
Output Choices • Advanced
¦ Species Fractions Options
¦ Species Fractions - Advanced
PM2.5 Calculation Options
¦ Model Data Options
PM2.5 Monitor Date Years
Start Year 12000 [~] End Year [ 2004
"31
® Official Desiqn Values
O Custom Desiqn Values
Valid FRM Quarters
Minimum days for valid quarter
Valid FRM Design Values
Minimum valid quarters in design value period |
Valid FRM Monitors
Minimum Number of Design Values
Required Design Values
None selected
3]
NH4 future calculation
(* Calculate future year NH4 using base year (constant) DON values
C Calculate future year N H 4 using base year N H 4 and the N H 4 R R F
< Back
Next >
Cancel
You also can use the Model Data Options to specify how to use the model data. This is
described in the Model Data Options section of the Annual PM Analysis: Details chapter.
Abt Associates Inc.
34
June 2010
-------
Overview of MATS Components
The last step is to verify the inputs to the analysis.
Annual PM Analysis
Choose Desired Output
Output Choices - Advanced
Final Output Choices and Verification
Verify inputs
¦¦Prsss.heret^
Checking...
Check OK. Press the finish button to continue..
Save Scenario
< Back Save Scenario & Run Cancel
3.1.2 Daily PM Analysis
With the Standard Analysis, MATS can forecast daily PM2.5 design values at monitor
locations. MATS can also calculate quarterly model data files and a species fractions file.
The Choose Desired Output window lets you specify the type of calculation(s) that you
would like MATS to perform. These different assumptions are discussed in the Output
Choice section of the Daily PM Analysis: Details chapter.
Abt Associates Inc.
35
June 2010
-------
Overview of MATS Components
Daily PM Analysis
Choose Desired Output
¦ Output Choices-Advanced
¦ Data Input
¦ Species Fractions Options
¦ Species Fractions - Advanced
¦ PM 2.5 Calculation 0 ptions
¦ Model Data Options
¦ Final Output and Check
Choose Desired Output
Scenario Name: [Example Daily PM
Standard Analysis
W Interpolate speciation monitor data to FRM monitor sites. Temporally-adjust.
Quarterly Peak Model Data
|s? Output quarterly peak model data file.
Species Fraction
p iOutput species fractions file.
Actions on run completion
p Automatically extract all selected output files
In the Output Choice Advanced window, MATS lets you choose from among a variety
of options that are generally used for quality assurance (QA). Details regarding these
choices are in the Output Choice - Advanced section of the Daily PM Analysis: Details
chapter.
Abt Associates Inc.
36
June 2010
-------
Overview of MATS Components
Daily PM Analysis
Choose Desired Output
Output Choices - Advanced
¦ Data Input
¦ Species Fractions Options
¦ Species Fractions - Advanced
¦ PM 2.5 Calculation 0 ptions
¦ Model Data Options
¦ Final Output and Check
Output Choice - Advanced
Miscellaneous Outputs
Quarterly average files
R jPoinll
High county sites
W File "C"
Quarterly average speciated monitors
|7 File "E"
Neighborfiles
p Point
>
< Back
Next >
Cancel
In the Data Input window, you specify the MATS input files that are used in each scenario.
There are three main types of files which must be specified. These include ambient
PM2.5 species data, ambient total PM2.5 data (FRM and IMPROVE), and gridded model
output data (e.g. CMAQ or CAMx data). There is specific terminology that is used on the
Data Input page. "Official" data refers to PM2.5 FRM data that can be used to determine
official design values for compliance purposes (comparison to the NAAQS). Other
datasets which may not have rigid regulatory significance are sometimes referred to as
"unofficial" data. The format for the data is in the Data Input section of the Daily PM
Analysis: Details chapter.
Abt Associates Inc.
37
June 2010
-------
Overview of MATS Components
Daily PM Analysis
Choose Desired Output
¦ Output Choices-Advanced
Data Input
¦ Species Fractions Options
¦ Species Fractions - Advanced
¦ PM 2.5 Calculation 0 ptions
¦ Model Data Options
¦ Final Output and Check
Data Input
Species Data
© ISpecies Monitor Data File
|C:\Prograrn Files'^Abt Associates\MATS\S ampleD ata\Species-f •"!
|p0H J |
d
| spatial field
d
PM2.5 Monitor Data
Unofficial Daily Average PM2.5 Data File fforAII Species Fractions & PM2.5 Spatial Field1)
|C: \Program Files\Abt Associates\MAT S \S ampleD ata\PM 25-for-fractions-0206-v2. csv
Official Daily Average FRM Data File (forPM2.5 Point Calculations)
| C: \Program FilesSAbt AssociatesSM AT S \S ampleD ata\24-hr-official-FR M -99-07-v2. csv
Model Data
© Daily model data input O Quarterly peak model data input
Baseline File |C:\Program FilesSAbt Associates\MATS\SampleData\2002cc_EUS_PM25_sub.csv
Forecast File | C: \Program FilesSAbt Associates\MAT S \S ampleD ata\2020cc_E U S_PM 25_sub. csv ~\
Location Helper |
The Species Fractions Calculation Options has two main sections. One involving
speciated monitor data (e.g., STN and IMPROVE monitors) and the other total PM2.5
monitor data (FRM and IMPROVE). For each type of data you can specify the years of
interest, whether you want to delete certain data, and the minimum amount of data for a
monitor to be considered "valid" (and thus included in the calculations). Details on these
options are in theSpecies Fractions Calculation Options section of the Daily PM Analysis:
Details chapter.
Abt Associates Inc.
38
June 2010
-------
Overview of MATS Components
Daily PM Analysis
Choose Desired Output
¦ Output Choices - Advanced
¦ Data Input
Species Fractions Options
¦ Species Fractions - Advanced
¦ Model Data Options
¦Kl Final Output and Check
Species Fractions Calculation Options
^3
IMPROVE-STN Monitor Data
Monitor Data Years
Start Year End Year
Delete Specified Data Values
Jv EPA-specified deletions from monitor data
User-specified deletions from monitor data
Minimum Data Requirements
Minimum number of valid days per quarter
Minimum number of valid quarters per valid year
Minimum number of valid years required for valid monitor
PM2.5 Monitor Data
Monitor Data Years
Start Year End Year
12002 jd [2004 3
Delete Specified Data Values
[~ EPA-specified deletions from monitor data
User-specified deletions from monitor data
Minimum Data Requirements
Minimum number of valid days per quarter | 11
Minimum number of valid quarters per valid year (point calculations)
Minimum number of valid years required for valid monitor | 1 -y-|
< Back
Next >
Cancel
The Species Fractions Calculation Options - Advanced screen allows you to make
relatively advanced choices for your analysis. Generally speaking, the default options
settings are consistent with the EPA modeling guidance document (note: the start and end
years should always be set to match the relevant base modeling year). A first set of options
provides different options for choosing peak monitor days. A second set of options allows
you to specify the interpolation weighting that you want to use and whether the
interpolation involves a maximum distance or not. The third set of options involves
choices regarding ammonium, blank mass, and organic carbon. Details on these options
are in the Species Fractions Calculation Options - Advanced section of the Daily PM
Analysis: Details chapter.
Abt Associates Inc.
39
June 2010
-------
Overview of MATS Components
Daily PM Analysis
^^Ispecies Fractions - Advanced
¦ PM2.5 Calculation Options
¦ Model Data Options
I Final Output and Check
Species Fractions Calculation Options - Advanced
Using Monitor Data to Calculate Species Fractions
IMPROVE-STN Monitor Data
® iUse top X percent of daily monitor days! | 10-$~|
Use all daily monitor values greater than i ri
fixed amount (uq/m3)
Minimum number of days required above fixed amount 11
O Use top X number of daily monitor days | 25-^-j
PM2.5 Monitor Data
<•) Use top X percent of daily monitor days | 10~ri
Use all daily monitor values greater than i jj-ri
~ fixed amount (uq/m3) ' ^
Minimum number of days required above fixed amount j 1 ~r|
Q Use top X number of daily monitor days | 25-—¦]
Interpolation Options
PM2.5 |~ 90000j< Crustal | Inverse Distance Squared V ~ 190000j< E:».
S04 | Inverse Distance Squared V ~] [90000j« DON
N03 | Inverse Distance Squared V ~] 190000 *}~7"| »\ OC
EC | Inverse Distance Squared V ~ ] |90000 **| ^H4
Salt | Inverse Distance Squared V ~ ] 190000 »|
Miscellaneous Options
Ammonium
® Use DON values
Use measured ammonium
NH4 percentage evaporating (0-100)
| Inverse Distance Squared V ~ 1 |90000 »|
| Inverse Distance Squared V ~ [ |90000 »|
nverse Distance Squared J l »: 0 H I
03
Default Blank Mass
Default Blank Mass
Organic Carbon
Organic carbon mass balance floor f
Organic carbon mass balance ceiling
0.5±|
~r±l
0.G^j
< Back
Next >
Cancel
The PM2.5 Calculation Options window allows you to specify the particular years of
monitor data that you want to use from the input file you specified in the Data Input
section. You can specify whether monitors should have a minimum number of design
values or a design value for a particular year. You can also specify how to calculate future
NH4 levels. Details on these options are in the PM2.5 Calculation Options section of the
Daily PM Analysis: Details chapter.
Abt Associates Inc.
40
June 2010
-------
Overview of MATS Components
Daily PM Analysis
Choose Desired Output
PM2.5 Calculation Options
¦ Species Fractions Options
¦ Species Fractions • Advanced
JPM2.5 Calculation Options
PM2.5 Monitor Data Years
Start Year |2000
¦^1 End Year 12004
T3
Valid FRM Monitors
Minimum Number of Design Value Periods fl ^
R equired D esign Value Periods |None selected
"3]
NH4 future calculation
(* Calculate future year NH4 using base year (constant) DON values
r Calculate future year NH4 using base year NH4 and the NH4 RRF
You also can use the Model Data Options to specify how to use the model data. This is
described in the Model Data Options section of the Daily PM Analysis: Details chapter.
Abt Associates Inc.
41
June 2010
-------
Overview of MATS Components
The last step is to verify the inputs to the analysis.
3.1.3 Ozone Analysis
MATS can forecast ozone design values at monitor locations — these forecasts are referred
to as Point Estimates. MATS can also use a variety of approaches to calculate design
values for a Spatial Field. The Choose Desired Output window lets you specify the type
of calculation(s) that you would like MATS to perform. These different assumptions are
discussed in the Choose Desired Output section of the Ozone Analysis: Details chapter.
Abt Associates Inc.
42
June 2010
-------
Overview of MATS Components
The Data Input window lets you specify the data files that you want to use. MATS comes
populated with default data sets, but you can use your own data if you choose. The format
for the data is in the Data Input section of the Ozone Analysis: Details chapter.
The Data Input window also lets you choose how to use model data when calculating a
temporal adjustment at a monitor. This is discussed in detail in the Using Model Data
section of the Ozone Analysis: Details chapter.
Abt Associates Inc.
43
June 2010
-------
Overview of MATS Components
B
Data Input
Monitor Data
~zone Data |mpleData\0ZQNE_MATS_input_99Q7-v2.csv ¦ ¦¦ |
Model Data
Baseline File fs~\SannpleData\ozone_mQdel_data_2QQ1.csv • |
Forecast File [s\S annpleD ata\ozone_rinQdel_data_2Q15.csv ¦"!
Using Model Data
T ernporal adjustment at monitor | 3x3 ^ | [Maximum
< Back
Next >
Cancel
The Filtering and Interpolation window lets you specify the years of data that you want
to use, any restrictions you want to apply when choosing valid monitors (i.e., monitors that
MATS use in its calculations), and options on the interpolation method. This is discussed
in detail in the Filtering and Interpolation section of the Ozone Analysis: Details chapter.
Abt Associates Inc.
44
June 2010
-------
Overview of MATS Components
Desired output
Data Input
Filtering/Interpolation
RRF/Spatial Gradient
Filtering and Interpolation
Choose Ozone Design Values
Start Year
| -r End Year 12002-2004 T]
Valid Ozone Monitors
Minimum Number of design values [T~
Max Distance from Domain [km] 25
Required Design Values [None selected
~3
Default Interpolation Method
| Inverse Distance Weights
r~ check to set a maximum interpolation distance [km]
I
[Too
< Back
Next >
Cancel
The RRF and Spatial Gradient window lets you set parameters used in the calculation of
relative response factors fRRF) and spatial gradients. This is discussed in detail in the
RRF and Spatial Gradient section of the Ozone Analysis: Details chapter.
Abt Associates Inc.
45
June 2010
-------
Overview of MATS Components
!¦[. Desired output
¦ Data Input
9 Filtering/Interpolation
RRF/Spatial Gradient
¦ Final Check
RRF and Spatial Gradient
S3
RRF Setup:
Initial threshold value (ppb)
Minimum number of days in baseline at or above threshold
Minimum allowable threshold value (ppb)
Min number of days at or above minimum allowable threshold [
f Enable Backstop minimum threshold for spatial fields
Backstop minimum threshold for spatial fields
Subrange first day of ozone season used in RRF
Subrange last day of ozone season used in RRF
r Pair days based on high concentration instead of date.
Spatial Gradient Setup:
Start Value
End Value
|1
1153
35
10
70
< Back
Next >
Cancel
The last step is to verify the inputs to the analysis.
Desired output
Data Input
Filtering/I nterpolation
RRF/Spatial Gradient
Final Check
Final Check
Verify inputs
Press here to verify your selections...
Save Scenario
< Back Cancel
Abt Associates Inc.
46
June 2010
-------
Overview of MATS Components
3.1.4 Visibility Analysis
MATS can forecast visibility in Class I Areas - these forecasts are referred to as Point
Estimates. In addition to specifying the Scenario Name, you can choose the version of the
IMPROVE Algorithm that you want to use. You can also choose whether to use model
data at the monitor linked to each Class I Area, or whether to use model data closest to the
Class I Area centroid. These different assumption are discussed in the Desired Output
section of the Visibility Analysis: Details chapter.
Choose Desired Output
Final Check
Point Estimates
Scenario Name : |EKannple Visibility
Forecast
|7 T emporally-adjust visibility levels at Class 1 Areas
IMPROVE Algorithm
C use old version
(• use new version
© Use model grid cells at monitor
Use model grid cells at Class 1 area centroid
Actions on run completion
Automatically extract all selected output files!
< Back
Next >
Cancel
The Data Input window lets you specify the data files that you want to use. MATS comes
populated with default input data, but you can use your own data if you choose. The
format for the data is in the Data Input section of the Visibility Analysis: Details chapter.
The Data Input window also lets you choose how to use model data when calculating a
temporal adjustment at a monitor. This is discussed in detail in the Using Model Data
section of the Visibility Analysis: Details chapter.
Abt Associates Inc.
47
June 2010
-------
Overview of MATS Components
¦ Choose Desired Output
Data Input
¦ Filtering
ESI Final Check
Data Input
Monitor Data
IMPROVE Monitor Data • Old Algorithm JCAProgrann Files^Abt Associates\MATS\Sarnp ¦¦¦
IMPROVE Monitor Data ¦ New Algorithm |006-daily IMPROVE-all data-new equation.csv| ¦¦¦!
Model Data
Baseline File
Forecast File
Using Model Data
T emporal adjustment at monitor
|S\SampleData\2002cc_EUS_PM25_sub.csv
|S\SampleData\2020cc_ELIS_PM25_sub.csv
3x3 "r
< Back
Next >
Cancel
The Filtering window lets you specify the years of data that you want to use, and any
restrictions you want to apply when choosing valid monitors (i.e., monitors that MATS use
in its calculations). This is discussed in detail in the Filtering section of the Visibility
Analysis: Details chapter.
Abt Associates Inc.
48
June 2010
-------
Overview of MATS Components
¦ Choose Desired Output
¦ Data Input
Filtering
ESI Final Check
Filtering
Choose Visibility Data Years
Start Monitor Year End Monitor Year Base Model Year
ISNHflQ 12004 T] 12002 V]
Valid Visibility Monitors
Minimum years required for a valid monitor 3
Man Distance from Centroid to Gridcell Center [km] 25
< Back
Next >
Cancel
The last step is to verify the inputs to the analysis.
Abt Associates Inc.
49
June 2010
-------
Overview of MATS Components
Choose Desired Output
Data Input
Filtering
Final Check
Final Check
Verify inputs
Save Scenario
< Back Save Scenario & Run Cancel
3.2 Output Navigator
The Output Navigator allows you to load results files (i.e., ASR files) that you have
previously created in MATS. You can view these data in maps and in tables, or export the
data to text files that you can then work with in a program such as Excel.
To start, just click on the Output Navigator tab. Then click on the Load button to choose
the file that you want to examine. You can click the Extract All button, and MATS will
create a folder with all of the files that MATS has generated. (A default name for the
folder is the Scenario Name you have chosen.)
Extracting ALL...
The files generated by MATS are of two types: (1) Configuration and Log files; and (2)
Output files containing the results of the MATS calculations.
Abt Associates Inc.
50
June 2010
-------
Overview of MATS Components
C:\Program FileslAbt Associates\MATS\output\Example 03
File Edit View Favorites Tools Help
V I
QBack * Q " yP5earch
i; Folders
Hi-
Address |ir^ C:\Program Files\Abt Associates\MAT5\output\Exarnple 03
Folders '
B-O MAT5
H£j configs
Q data
Q help
i (£) maps
&£| output
._j Example Visibility
; Q Tutorial 03
Tutorial Visibility
IB l£| sampledata
IS work
Name
\m Configuration, cfg
^Example 03 - Ozone Monitors ~ county high monitoring sites, temporally adjusted 2015,csv
^Example 03 - Ozone Monitors ~ monitor data, temporally adjusted 2015,csv
^Example 03 - Spatial Field — interpolated monitor data, temporally adjusted; gradient-adjusted monitor data, 201!
0 Log File.log
Example 03
Another option is to right-click on a particular file, and then you can choose whether to use
data to Add to Map, View, or Extract.
B0H
Help
Start Map View I Output Navigator I
Load
Extract All
Highlight file of interest and right-click to view options to Map, View, and Extract the data.
I Type
Name
- Configuration/Log Files
Configuration
Log File
- Output Files
Example 03 - Ozone Monitors - monitor
Example 03 - Ozone Monitors - county h ¦
Example 03 - Spatial Field - interpolate!
djusted 2015
Configuration
Run Log
Monitor Network
mmim
| es, temporally adjusted 201 Monitor Network
mporally adjusted; gradient- Monitor Network
81 kb
1 kb
56kb
3464kb
J | Stop
The View option lets you examine the data and then to export it to a CSV file, which you
can then load into another program such as Microsoft Excel.
Abt Associates Inc.
51
June 2010
-------
Overview of MATS Components
Help "
Start Map View Output Navigator | Monitor Network Data |
Example 03 - Ozone Monitors — monitor data, temporally adjusted 2015
Show All or select a particular location to see data.
id
~\w
[long
010030010
010270001
010331002
010510001
010550011
010690004
010730023
ni n-701 nni
30.498001
33.281261
34.760556
32.4985667
33.904039
31.1906565
33.553056
AOCCCC
-87.8814123
-85.8021817
-87.650556
-86.1365871
-86.0538672
-85.423117
-86.815
ojLHL
Is
Select Quantities that must be >= 0
~ b_o3_dv
~ f_o3_dv
~ referencecell
J rrf
~ ppb
~ days
Export Exp o rt cu rre ntly s h own d ata to CSV
id
date
|b_o3_dv |f_o3_dv |referencece|rrf | ppb [days ~|
010030010
010270001
010331002
010510001
010550011
010690004
mn7^nn?3
2004
2004
2004
2004
2004
2004
?nri4
78.0
79.3
-7.00
68.8
62.7
-9.00
95023
108051
92063
76.7
63.2
106043
75.0
58.0
105056
-7.00
-9.00
113032
77,Q
6n?
mnnR?
0.8825
0.7909
0.7642
0.8251
0.7738
-9.00
0 7828
85.0
71.0
71.0
70.0
73.0
70.0
_£UL
11.0
11.0
11.0
9.00
10.0
1.00
11 n
_a
Stop Info
Choosing the Extract option will allow you to immediately export the data to a CSV file.
The default file name for the CSV file is the same one that you see in the Output Navigator
window {e.g., Example 03 - Ozone Monitors — monitor data.csv). Finally, choosing the
Add to Map option allows you to create a map of your results.
3.3 Map View
The Map View allows you to perform a variety of mapping tasks. You can zoom in to a
particular location; choose particular colors to map your data, export the maps you have
created to BMP files, among other things. These various options are discussed in detail in
the Map View chapter.
Abt Associates Inc.
52
June 2010
-------
Overview of MATS Components
Start | Map View | Output Navigator
J | Stop Info
" +, "... PSy (JJj Standard Layers*•
JiData Loaded ||
@ ° Example 03-Ozone Monitor;
© • f_o3_dv-9to-9
0 • f_o3_dv-9 to 68.1
® © f_o3_dv 68.1 to 72.2
© O f_o3_dv 72.2 to 77.2
© O f_o3_dv 77.2 to 96.3
Long:-94.77210, Lat: 52.46175 ***
Extent: Min(-113.747,23.295) Max(*H.284,41.094)
3.4 Help
The Help dropdown menu has the User Manual for MATS and version information.
Abt Associates Inc.
53
June 2010
-------
Overview of MATS Components
Help
About MATS
User Manual
Navigator
uzone Anaiys
Visibility Analysis
Annual PM Analysis
Daily PM Analysis
Stop Info
Abt Associates Inc.
54
June 2010
-------
Annual PM Analysis: Quick Start Tutorial
4 Annual PM Analysis: Quick Start Tutorial
In this tutorial you will forecast annual PM2.5 design values at monitors in the Eastern
United States. The steps in this analysis are as follows:
Step 1. Start MATS. Start the MATS program and choose to do an Annual PM
analysis.
Step 2. Output Choice. Choose the output to generate. In this example, you will do two
things: forecast annual PM2.5 levels at monitor locations and output a species fractions
file (which you can subsequently reuse, as discussed here!
Step 3. Output Choice - Advanced. With these advanced options, you can generate
spatial fields and a variety of files useful for quality assurance. Simply review these
options and then uncheck them all. (If you are interested, these options are all described
here.')
Step 4. Data Input. Choose the particular years of data and monitors to use in this
analysis.
Step 5. Species Fractions Calculation Options. Specify how to generate the relative
response factors fRRFs) used in the forecasts.
Step 6. Species Fractions Calculation Options - Advanced. This window allows you to
make relatively advanced choices for your analysis, such as choosing different ways to
interpolate the monitor data.
Step 7. PM2.5 Calculation Options. Among other things you can specify the particular
years of monitor data that you want to use.
Step 8. Model Data Options. Choose how to use the model data, such as determining
the maximum distance the model data can be from a monitor.
Step 9. Final Check. Verify the choices you have made.
Step 10. Map Output. Prepare maps of your forecasts.
Step 11. View & Export Output. Examine the data in a table format.
Each step is explained below. Additional details are provided in the section Annual PM
Analysis: Details.
4.1 Stepl. Start MATS
Double-click on the MATS icon on your desktop, and the following window will appear:
Abt Associates Inc.
55
June 2010
-------
Annual PM Analysis: Quick Start Tutorial
Umats
Help T
Start Map View Output Navigator
Ozone Analysis
Visibility Analysis
Annual PM Analysis
Daily PM Analysis
Stop Info
Click the Annual PM Analysis button on the main MATS window. This will bring up
the Configuration Management window.
Configuration Management
(* ICreate New Configuration!
r Open Existing Configuration
Go
Cancel
A Configuration allows you to keep track of the choices that you make when using MATS.
For example, after generating results in MATS, you can go back, change one of your
choices, rerun your analysis, and then see the impact of this change without having to enter
in all of your other choices. For this example, we will start with a New Configuration.
Choose Create New Configuration and click the Go button. This will bring up the
Choose Desired Output window.
4.2 Step 2. Output Choice
The Choose Desired Output window allows you to choose the output that you would like
Abt Associates Inc. June 2010
56
-------
Annual PM Analysis: Quick Start Tutorial
to generate. MATS allows you to conduct a Standard Analysis (i.e., forecast Point
Estimates at ambient monitors), output quarterly model data, and output a species fractions
file.
In the Scenario Name box type " Tutorial Annual PM' - this will be used to keep track
of where your results are stored and the variable names used in your results files.
Standard Analysis. Leave the box checked next to "Interpolate monitor data to FRM
monitor sites. Temporally-adjust." MATS will create forecasts for each monitor in the
monitor file. (Additional details are in the Standard Analysis section.)
Quarterly Model Data. Uncheck these options. (If checked, MATS generates quarterly
model files that MATS generates from daily data that you have provided. This is useful
if you want to reuse model files — the quarterly files are much smaller and MATS will
run faster if it can skip the step of creating quarterly data from the daily. These files are
described here.)
Species Fraction. Check the box next to Output species fractions file. This will
generate a reusable file described here.
Actions on run completion. Check the box next to Automatically extract all selected
output files. Upon completing its calculations, MATS will extract the results into a
folder with the name of your scenario.
Choose Desired Output
Scenario Name: |Tutorial Annual PM
Standard Analysis
p Interpolate monitor data to FRM monitor sites. Temporally-adjust.
Quarterly Model Data
Output quarterly average model data file.
Output used quarterly average model data file.
Species Fraction
Putput species fractions filej
Actions on run completion
Automatically extract all selected output files
Annual PM Analysis
f Choose Desired Output
¦ Output Choices - Advanced
¦ Data Input
¦ Species Fractions Options
¦ S pecies Fractions - Advanced
¦ PM2.5 Calculation Options
¦ Model Data Options
Final Output and Check
Next > Cancel
When your window looks like the window above, click Next. This will bring you to the
Output Choice - Advanced window.
Abt Associates Inc.
57
June 2010
-------
Annual PM Analysis: Quick Start Tutorial
4.3 Step 3. Output Choice - Advanced
With the advanced options in the Output Choice - Advanced window, you can generate
spatial fields and a variety of files useful for quality assurance. Simply review these
options and then uncheck them all. (If you are interested, these options are all described
here.)
Output Choice - Advanced
Spatial Field Estimates
Forecast
Interpolate FRM and speciation monitor data to spatial field. Temporally adjust.
r~ Interpolate gradient-adjusted FRM and speciation monitor data to spatial field. Temporally adjust.
Miscellaneous Outputs
Quarterly average files
Point Spatial Field
Spatial Field ¦ gradient-adjusted
High county sites
r File "C"
Species fractions spatial field
Spatial Field Spatial Field - gradient-adjusted
Quarterly average speciated monitors
I-File "E'i
Neighborfiles
Point V Spatial Field
< Back
Next >
Cancel
Annual PM Analysis
Choose Desired Output
Output Choices - Advanced
¦ Data Input
¦ Species Fractions Options
¦ Species Fractions - Advanced
¦ PM2.5 Calculation Options
¦ Model Data Options
When your window looks like the window above, click Next. This will bring you to the
Data Input window.
4.4 Step 4. Data Input
The Data Input window allows you to choose the species and PM2.5 monitor data and the
model data that you want to use. As discussed in more detail in the following chapter (see
Standard Analysis). MATS calculates the ratio of the base and future year model data to
calculate a relative response factor (RRF) for each PM species. MATS uses the PM2.5
monitor data and interpolated species monitor data to estimate species values at each FRM
site, multiplies the species values from the monitor data with the species-specific RRFs,
and then estimates a future-year design value. (Additional details on Data Input are
available here.)
Use the default settings in the Data Input window. The window should look like the
following:
Abt Associates Inc. June 2010
58
-------
Annual PM Analysis: Quick Start Tutorial
Annual PM Analysis
Choose Desired Output
¦ Output Choices - Advanced
Input
¦ Species Fractions Options
¦ Species Fractions - Advanced
¦ PM2.5 Calculation Options
¦ Model Data Options
BEI Final Output and Check
Species Data
(• Species Monitor Data File
O Species Fractions File [
F
|sociates\MATS\SarnpleData\Species-f or-f ractions-0206-v2.csv • |
"^r
~3
PM2.5 Monitor Data
Unofficial Daily Average PM2.5 Data File ffor All Species Fractions & PM2.5 Spatial Fieldl
[C: \Prograrn FilesSAbt Associates\MAT S \S ampleD ata\PM 25-for-fractions-020G-v2. csv
Official Quarterly Average FRM Data File (for PM2.5 Point Calculations)
[C: \Prograrn FilesSAbt Associates\MAT S \S ampleD ataSAnnual-official-FR M -99-07-v2. csv
Model Data
© Daily model data input O Quarterly model data input
Baseline File |C:\Program FilesSAbt Associates\MATS\SampleData\2002cc_EUS_PM25_sub.csv
Forecast File | C: \Program FilesSAbt Associates\MAT S \S ampleD ata\2020cc_E U S_PM 25_sub. csv
"3
IB
"3
< Back
Next >
Cancel
When your window looks like the window above, click Next. This will bring you to the
Species Fractions Calculation Options window.
4.5 Step 5. Species Fractions Calculation Options
The Species Fractions Calculation Options window has several functions related to the
IMPROVE-STN (species) monitor data and the (unofficial) PM2.5 monitor data. These
functions include identifying the years of monitor data that you want to use, deleting any
specific data values, and choosing the minimum data requirements of monitors you want in
your analysis.
Monitor Data Years. Choose the years of monitor data that you want to use. The
default is to use the three-year period 2002-2004. (That is, for both IMPROVE-STN and
PM2.5 monitor data, the Start Year is 2002 and the End Year is 2004.)
Delete Specified Data Values. The default is to delete the observations specified by
EPA. As described in the Data Input section, valid data are given a value of "0" and
observations that should be deleted are given a value of" 1" to "10". (Leave unchecked
the option for the user to flag data.)
Minimum Data Requirements. There are three sets of minimum data requirements:
1. Minimum number of valid days per valid quarter. This is the minimum number
of site-days per valid quarter. The default is 11 days, which corresponds to >
Abt Associates Inc.
59
June 2010
-------
Annual PM Analysis: Quick Start Tutorial
75% completeness for monitors on a 1 in 6 day schedule. This is a minimum
number of samples that is routinely used in calculations of quarterly average
concentrations.
2. Minimum number of valid quarters required for valid season. This number of
years of data (within the start year and end year specified) for which we have
valid quarters for a given season. The default value is 1 year. If the value is set
= 2, then there will need to be 2 years of valid data from quarterl in order for
quarter one to be considered complete (and the same for the other 3 quarters).
3. Minimum number of valid seasons required for valid monitor. This is the
number of valid seasons that are needed in order for a particular monitor's data
to be considered valid. The default is 1 for IMPROVE-STN monitor data and
the range is 1-4. For example, if the value is = 1, then a monitor's data will be
used in the species fractions calculations if it has at least one valid season. If
the value = 4, then the site must have all 4 seasons of valid data to be used. The
default for PM2.5 depends on whether the data are used in point calculations
(default = 4) or spatial field calculations (default =1).
Use the default settings pictured in the screenshot below. (All of these options are
described in detail here.')
Abt Associates Inc.
60
June 2010
-------
Annual PM Analysis: Quick Start Tutorial
Annual PM Analysis
BfiJMChoose Desired Output
¦ Q utput Choices - Advanced
¦ Data Input
Species Fractions Options
¦ Species Fractions • Advanced
¦ PM2.5Calculation0ptions
¦ Model Data Options
l|^^|Final Output and Check
When your window looks like the window above, click Next. This will bring you to the
Species Fractions Calculation Options window.
4.6 Step 6. Species Fractions Calculation Options -
Advanced
The Species Fractions Calculation Options - Advanced screen allows you to make
relatively advanced choices for your analysis. Generally speaking, the default options
settings are consistent with the EPA modeling guidance document. One set of options
allows you to specify the interpolation weighting that you want to use and whether the
interpolation involves a maximum distance or not. The second set of options involves
choices regarding ammonium, blank mass, and organic carbon.
Use the default settings pictured in the screenshot below. (All of these options are
described in detail here.)
Abt Associates Inc.
61
June 2010
-------
Annual PM Analysis: Quick Start Tutorial
Annual PM Analysis
¦¦SBChoose Desired Output
¦ Output Choices • Advanced
¦ Data Input
¦ Species Fractions Options
Species Fractions - Advanced
¦ PM2.5 Calculation Options
¦ Model Data Options
Final Output and Check
When your window looks like the window above, click Next. This will bring you to the
PM2.5 Calculation Options window.
4.7 Step 7. PM2.5 Calculation Options
The PM2.5 Calculation Options window allows you to specify the particular years of
monitor data that you want to use from the input file you specified in Step 4 (Data Input).
Keep the default settings:
PM2.5 Monitor Data Years. Leave Start Year = 2000 and End Year = 2004.
Official vs. Custom Values. Specify "official" design values, which is the recommended
default setting.
Valid FRM Monitors. Keep the minimum number of design values equal to the default
value of 1, and do not specify any particular design values for inclusion in the
calculations.
NH4 Future Calculation. You can also specify how you want to forecast NH4 values.
Use the default approach, which is to use baseline DON values.
Use the default settings pictured in the screenshot below. (All of these options are
described in detail here.)
Abt Associates Inc.
62
June 2010
-------
Annual PM Analysis: Quick Start Tutorial
Annual PM Analysis
¦¦SBChoose Desired Output
¦ Output Choices • Advanced
¦ Data Input
¦ Species Fractions Options
¦ Species Fractions • Advanced
^JpM2.5 Calculation Options
¦ Model Data Options
Final Output and Check
PM2.5 Calculation Options
PM2 5 Monitor Data Years
tart Year [ 2000 * EndYear 12004
|ii"
® Official Desiqn Values
O Custom Desiqn Values
Valid FRM Quarters
Minimum days for valid quarter
Valid FRM Design Values
Minimum valid quarters in design value period |12
Valid FRM Monitors
Minimum Number of Design Values
Required Design Values
| None selected
_NH4 future calculation
(• Calculate future year NH4 using base year (constant) DON values
C Calculate future year NH4 using base year NH4 and the NH4 RRF
When your window looks like the window above, click Next. This will bring you to the
Model Data Options window.
4.8 Step 8. Model Data Options
The Model Data Options section allows you to specify:
Max Distance to Domain. This is the maximum distance from a given monitor to the
nearest model grid cell (measured in kilometers). If a monitor is further than the
specified maximum distance from the center of any grid cell, then MATS will not
calculate results for that monitor or use the monitor in any calculations (this effectively
eliminates monitors from outside the specified domain from the being included in the
output files).
Temporal Adjustment at Monitor. This option specifies how many model grid cells to
use in the calculation of RRFs for point estimates and for spatial estimates. Use the
default option: 3x3 set of grid cells. Note that for PM analyses, MATS calculates mean
concentrations across the grid cell array (as compared to maximum concentrations used
for ozone analyses).
Use the default settings pictured in the screenshot below. (All of these options are
described further here.)
Abt Associates Inc.
63
June 2010
-------
Annual PM Analysis: Quick Start Tutorial
When your window looks like the window above, click Next. This will bring you to the
Final Check window.
4.9 Step 9. Final Check
The Final Check window verifies the choices that you have made. For example, it makes
sure that the paths specified to each of the files used in your Configuration are valid.
Click on the Press here to verify selections button.
Abt Associates Inc.
64
June 2010
-------
Annual PM Analysis: Quick Start Tutorial
Annual PM Analysis
Choose Desired Output
¦ Output Choices - Advanced
¦ Data Input
¦ Species Fractions Options
B Species Fractions - Advanced
¦ PM 2.5 Calculation 0 ptions
¦ Model Data Options
Final Output and Check
Final Output Choices and Verification
Verify inputs
| Press here to verify your selections^
Checking...
Check OK. Press the finish button to continue..
Save Scenario < Back Save Scenario & Run Cancel
If you encounter any errors, go back to the choices you have previously made by clicking
on the appropriate part (e.g., Data Input) of the tree in the left panel, and then make any
changes required.
When your window looks like the window above, click either Save Scenario & Run or
Save Scenario. Save Scenario & Run will cause MATS to immediately run the scenario.
A temporary, new Running tab will appear (in addition to the Start, Map View and
Output Navigator tabs).
Abt Associates Inc.
65
June 2010
-------
Annual PM Analysis: Quick Start Tutorial
MATS
E0E
Help "
I Start Map View Output Navigator
Running |
I
Close ]
Name
| Last Message
Tutorial Annual PM.asr
Loading/creating model baseline data.
When the calculations are complete, a small window indicating the results are Done will
appear. Click OK.
Done
Mill
After clicking OK, MATS will open a folder with the results files already exported.
Abt Associates Inc.
66
June 2010
-------
Annual PM Analysis: Quick Start Tutorial
to C:\Program Files\Abt Associates\MATSVoutput\Tutorial Annual PM |_ |n| X|
File Edit View Favorites
T ools Help
©^ck Q ¦ ,t
Search Folders
m-
I Address li i C:\Prooram Files!Abt Associates'iMATS'ioutput'iTutonal Annual PM v
og°
Name
Configuration, cfq
Log File,log
^Tutorial Annual PM Annual PM25 Point, csv
^Tutorial Annual PM Quarterly Avg Spec Frac Point.csv
Size Type
224 KB CFG File
2KB Text Document
46 KB Microsoft Excel Comma
137 KB Microsoft Excel Comma S
v <
Output Navigator tab will also be active. (The Running tab will no longer be seen.)
MATS will automatically load the output files associated with the .asr configuration that
just finished running.
Abt Associates Inc.
67
June 2010
-------
Annual PM Analysis: Quick Start Tutorial
Help
Start Map View | Output Navigator I
Load
Extract All
Highlight file of interest and right-click to view options to Map, View, and Extract the data.
| Type
Name
Size
B Configuration/Log Files
j-Configuration
Log File
B Output Files
j -Tutorial Annual PM Quarterly Avg Spec Frac Point
Tutorial Annual PM Annual PM25 Point
Configuration
Run Log
Monitor Network
Monitor Network
223kb
1kb
136kb
45kb
Stop Info
The next step (click here) shows you how to map your results with the Output Navigator.
For more details on mapping and other aspects of the Output Navigator, there is a
separate chapter on the Output Navigator.
4.10 Step 10. Map Output
After generating your results, Output Navigator can be used to load and/or map them. If a
run just finished, the output files will already be loaded into output navigator.
If files from a previous run need to be loaded then click on the Load button and choose
the Tutorial Annual PM.asr file.
Abt Associates Inc.
68
June 2010
-------
Annual PM Analysis: Quick Start Tutorial
Open MATS Result file
My Recent
Documents
r3
Desktop
My Documents
ail
My Computer
S!
My Network
Places
Look in: output
l£)Example 03
_jExannple Visibility
j Tutorial Annual PM
j Tutorial 03
j Tutorial Visibility
Example 03.asr
^Example Visibility.asr
Tutorial Annual PM.asr
] Tutorial 03.asr
] Tutorial Visibility,asr
3
S &
File name:
Files of type:
Tutorial Annual PM.asr
MATS Result File
~3
"3
Open
Cancel
Under Configuration/Log Files, you will see two files:
Configuration: keeps track of the assumptions that you have made in your analysis.
Log File : provides information on a variety of technical aspects regarding how a results
file (*.ASR) was created.
Under Output Files you will see:
Tutorial Annual PM Quarterly Avg Spec Frac Point: contains species fractions and
interpolated species values. (Note that this is a reusable file that you can load into
MATS.)
Tutorial Annual PM Annual PM25 Point: contains base & future PM2.5, species values,
and RRFs. (Note that the annual RRFs and annual species values are not used anywhere
in the calculation of design values, and here just for information.)
Right-click on the file Tutorial Annual PM Annual PM25 Point. This gives you three
options: Add to Map, View, and Extract. Choose the Add to Map option.
Abt Associates Inc.
69
June 2010
-------
Annual PM Analysis: Quick Start Tutorial
Help T
Start Map View | Output Navigator |
EHE
Load
Extract All
Highlight file of interest and right-click to view options to Map, View, and Extract the data.
|Type
Name
Size
B Configuration/Log Files
j-Configuration
Log File
& Output Files
Tutorial Annual PM Quarterly Avg Spec Frac Point
Tutorial Annual PM Annua
View
Extract
Configuration
Run Log
Monitor Network
223kb
1 kb
136kb
Stop Info
This will bring up the Map View tab.
Abt Associates Inc.
70
June 2010
-------
Annual PM Analysis: Quick Start Tutorial
MATS
Help
Start | Map View | Output Navigator
+ v' W"!) 'V" ® [3 Standard Layers
J; Data Loaded
© ° Tutorial Annual PM Quarterly A
Long:-179.16062, Lat: 63.29643 ***
Extent: Min(-159.661,2.334) Max(-H*44,51.197)
To view an enlarged map, use the Zoom to an area Task Bar button on the far left.
Choose the Continental US.
Abt Associates Inc.
71
June 2010
-------
Annual PM Analysis: Quick Start Tutorial
Continental US
V mats
~0®
Start | Map View | Output Navigator
Long:-159.36579, Lat: 13.88071 ^
Extent: Min(-156.233,3.410) Max(-8.833,54.39^
W \ \
Full Extent
v fit > Standard Layers T
Edit Zoom Frames
Add Current View to List
Maryland
New England
Southern California
Texas
Washington DC
Annual PM
To more easily view the location of monitors in particular states, uncheck US Counties
using the Standard Layers drop down menu on the far right of the Task Bar. Your
window should look like the following:
Abt Associates Inc.
72
June 2010
-------
Annual PM Analysis: Quick Start Tutorial
7 MATS
Start | Map View | Output Navigator
© # Tutorial Annual PM Annual
Long: -126.25521, Lat: 52.26646 ***
Extent: Min(-122.300,20.064) Max(-56.629,47*»"6)
Info
Zoom in further on the Eastern US using the Zoom in button on the Task Bar. This allows
you to view the results more closely. A dashed line surrounds the area that you have
chosen and should look something like the following:
Abt Associates Inc.
73
June 2010
-------
Annual PM Analysis: Quick Start Tutorial
Help ^
Start | Map View | Output Navigator
+, " +. , ^ Standard Layers T
JpataLoaded
@ • Tutorial Annual PM Annual PM
Long: -77.64010, Lat: 28.60642 ***
Extent: Min(-122.300,20.064) Max(-56.629,47w6)
|| 11 Stop Info
Right click on the "Tutorial Annual PM Annual PM25 Point" layer in the panel on the left
side of the window. Choose the Plot Value option
Abt Associates Inc.
74
June 2010
-------
Annual PM Analysis: Quick Start Tutorial
MATS
Help -
Start | Map View | Output Navigator
I,"1!) ^ T SS ^ Standard Layers -
I Data Loaded!
@ • Tutorial/
Remove
Export as CSV File
Long: -24.434%, Lat: 31.34245
Extent: Min(-94.739,30.465) Max(-69.763,38.
This will bring up Shape Class Breaks window. In the Value drop-down list, choose the
variable "f pin 25 aim dv" — this is forecasted PM2.5 design value.
Shape Class Breaks
Layer Name: Tutorial Annual PM Annual PM25 Point
Value:
Date
f_pm25_ann_dv
no date
ff Bins
"3
C Unique Values
Class Count: 5 ^ Marker Sizing: [c|
Start Color
End Color
Clear Breaks V' Apply X Close
Click Apply and then click Close. This will bring you back to the Map View window.
Abt Associates Inc.
75
June 2010
-------
Annual PM Analysis: Quick Start Tutorial
This is just a brief summary of the mapping possibilities available. For more details, there
is a separate chapter on the Map View. The next step is to go to the Output Navigator to
view the data in a table format.
4.11 Step 11. View Output
After mapping your results, click on the Output Navigator tab, so that you can then view
the data in a table. Right-click on the file Tutorial Annual PM Annual PM25 Point. This
gives you three options: Add to Map, View, and Extract. Choose the View option.
Abt Associates Inc.
76
June 2010
-------
Annual PM Analysis: Quick Start Tutorial
0@B
Help T
Start Map View | Output Navigator I
Load
Extract All
Highlight file of interest and right-click to view options to Map, View, and Extract the data.
|Type
Name
Size
B Configuration/Log Files
j- Configuration
Log File
B Output Files
Tutorial Annual PM Quarterly Avg SpecF.
Tutorial Annual PM Annual PM25 Point
Add To Map
Configuration
Run Log
Monitor Network
Monitor Network
223kb
1 kta
45kb
Stop InFo
This will bring up a Monitor Network Data tab. The upper left panel allows you to view
the ID and latitude and longitude of the monitors in your data — at the right of this panel
there is a scrollbar with which you can locate any particular monitor of interest.
Abt Associates Inc.
77
June 2010
-------
Annual PM Analysis: Quick Start Tutorial
Help
Start Map View Output Navigator | Monitor Network Data I
Tutorial Annual PM Quarterly Avg Spec Frac Point
Show All or select a particular location to see data.
id
~|typ
010270001
010331002
010491003
010550010
010730023
010731005
010731009
m moonno
33.281261
34.760556
34.287627
33.993749
[long
-85.802182
-87.650556
-85.968298
-85.991072
33.553056
33.331111
33.459722
-J-J .4on 7
-86.815
-87.305556 j
o i:¦ n a ..11 c 2 I
Select Quantities that must be >= 0
] pm25_mass_frac
] fcr
~ fee
~ fnh4
~ focm
~ fso4
~ fno3
~ fwater
~ fsalt
~ blank_mass
] don
~ i_so4
~ i no3r
_H|
Export Exportcurrentlyshown datato CSV
date |pm25_mass|fcr [fee
[fsalt |blank_mas;jdij'*'
id
fnh4
focm
fso4
fno3
fwater
010270001
010270001
010270001
010270001
10.65 0.0545
14.14 0.08011
17.15 0.07281
11.02 0.0554!
0.0691
0.0580
0.0438
0.0758
0.1083!
0.0995 i
0.0999!
0.0968;
0.3349
0.3189
0.3095
0.36111
JL2908I 0.0388
0.3476] (UI0080o|
0.3933| 0.00
0.29841 0.0121
010331002
010331002
-------
Annual PM Analysis: Quick Start Tutorial
Help ^
Start Map View 1 Output Navigator | Monitor Network Data |
Tutorial Annual PM Quarterly Avg Spec Frac Point
Close
Show All or select a particular location to see data.
~[vi
[long
010270001
33.281261
-85.802182
010331002
34.760556
-87.650556
010491003
34.287627
-85.968298
010550010
33.993749
-85.991072
010730023
33.553056
-86.815
010731005
33.331111
-87.003611
010731009
33.459722
-87.305556
nimT?nm I
tj rinniaal
QC mri1C7
sL
Select Quantities that must be >= 0
~ pm25_mass_frac
~ fcr
~ fee
~ fnh4
~ focm
~ fso4
~ fno3
~ fwater
~ fsalt
~ blank_mass
~ don
~ i_so 4
~ i no3r
M
Export Exp o rt cu rre ntly s h own d ata to CSV
id
date
pm25_mass
fcr Ifec
fnh4 |focm |fso4
fno3 (fwater
fsalt
blank_mass|don
010491003
Q1
11.68
0.0542 0.0529
0.131 0.255 0.3224
0.0737 0.1083
0.00240
0.500
010491003
Q2
15.42
0.0662 0.0379
0.110 0.3181 0.3494
0.000500 0.1169
0.000900
0.500
010491003
Q3
19.12
0.0536 0.0286
0.1068 0.3277 0.387
0.00 0.0955
0.000800
0.500
010491003
Q4
11.72
0.0514 0.0576
0.105 0.3554 0.303
0.0202 0.105
0.00250
0.500
E
-------
Annual PM Analysis: Details
5 Annual PM Analysis: Details
MATS can forecast annual design values at PM2.5 monitor locations — these forecasts are
referred to as Point Estimates. MATS can also use a variety of approaches to calculate
design values for a Spatial Field. A Spatial Field refers to a set of values comprising
calculations for each grid cell in a modeling domain from Eulerian grid models such as
CMAQ and CAMx.
The set of choices involved in calculating either Point Estimates or a Spatial Field can be
fairly involved, so MATS keeps track of these choices using a Configuration. When you
begin the process of generating PM2.5 estimates, MATS provides an option to start a new
Configuration or to open an existing Configuration.
Configuration Management
(5" iCreate New Configuration!
C Open Existing Configuration
Go
Cancel
Select your option and then click Go.
MATS will then step you through a series of windows with choices for your analysis.
Output Choice. Choose whether you want to run the Standard Analysis, and whether to
output a species fractions file and/or quarterly model data.
Output Choice - Advanced. This option provides miscellaneous Point Estimate output
files, as well as baseline and forecast year Spatial Fields and gradient-adjusted output.
Data Input. Load species monitoring data or a species fractions file. LoadPM2.5
ambient monitoring data. Finally, load the modeling data that you want to use.
Species Fractions Calculation Options. Choose the years of daily STN-IMPROVE and
FRM monitoring data. Identify valid monitors. Delete specified values.
Species Fractions Calculation Options - Advanced. Choose interpolation options for
PM2.5 and species monitoring data. Choose assumptions for the ammonium calculation,
default blank mass, and organic carbon.
PM2.5 Calculation Options - FRM Monitor Data. Choose the years of quarterly FRM
monitoring data. Choose whether to use official design values or custom design values.
Identify valid monitors. Choose the approach for calculating future NH4.
Model Data Options. Specify the maximum distance of monitors from modeling
domain. Specify which model grid cells will be used when calculating RRFs at monitor
locations.
Final Check. Verify the selections that you have made.
Abt Associates Inc.
80
June 2010
-------
Annual PM Analysis: Details
5.1 Output Choice
In the Output Choice window, MATS lets you specify the name of your Scenario, and
then to choose up to three options: Standard Analysis, which refer to forecasts made at
FRM PM2.5 monitor locations; Quarterly Model Data, which allows you to create
quarterly averages from daily model output data (output data from grid models such as
CMAQ and CAMx), and then subsequently reuse this file; and Species Fractions, which
outputs a reusable species fractions file.
By checking the box next to Automatically extract all selected output files, MATS will
create a separate folder with your chosen Scenario Name in the MATS "Output" folder,
and then export .CSV files with the results of your analysis. Alternatively, you can export
the results from the Output Navigator, but checking this box is a little easier.
Standard Analysis. The Standard Analysis refers to the calculation of future year
PM2.5 design values at FRM monitor locations. This is the main part of the modeled
attainment test for PM2.5. There are several calculations involved in this analysis. MATS
will interpolate PM2.5 species data, calculate species concentrations at each FRM site and
project design values to a future year using gridded model data. Most MATS users will
run this analysis and it is therefore checked by default.
Quarterly Model Data. MATS requires two types of data input: ambient monitor
data and gridded model output data. For the annual PM2.5 calculations, MATS will
accept either MATS formatted daily average gridded model files or quarterly average files.
If daily average model files are used as inputs, MATS will calculate quarterly averages
Abt Associates Inc.
81
June 2010
-------
Annual PM Analysis: Details
from the daily averages and optionally output the quarterly average concentrations into text
files (CSV files). The quarterly average text files can then be re-used in subsequent
MATS analyses. Quarterly average input files are smaller and run faster than daily
average files. There are two options to output quarterly average model concentration
CSV files:
Check the "Output quarterly average model data file" box to create quarterly average
CSV for all grid cells in the modeling domain. MATS will create one baseline year file
and one future year file. This will create relatively large files, but they will still be -90
times smaller than daily average files (assuming a full year of model data).
The second option is to check the "Output used quarterly average model data file". This
option will only output the grid cells that are subsequently used in the particular MATS
configuration. For example, if MATS calculates future year design values at 20 FRM
sites using a 1 X 1 grid array, then MATS will output base and future model values for
only 20 grid cells (assuming each monitor is in a unique grid cell). The advantage of
these files is that they are extremely small. But if subsequent MATS runs use a different
set of monitors or grid arrays, then the files may not contain all of the necessary data to
complete the analysis. This option is recommended as a QA tool to examine the grid
cells and the model concentrations that MATS is using in the analysis.
Species Fraction. Checking the "Output species fraction file" box will create an output
file containing the calculated PM2.5 species fractions at each FRM site used by MATS.
This species fraction file can be re-used in MATS as an input file. The species fraction file
can be useful for several reasons. One, using a species fraction file saves time because
MATS won't have to interpolate species data and calculate fractions each time it is run.
Two, it can provide consistency between MATS runs by ensuring that the same species
fractions are used each time. And for the same reason, the species fraction file can be used
interchangeably between different users to ensure that multiple groups are using the same
species fractions (if that is a goal). And finally, the fractions file can serve as a template
for creating a custom species fractions file using whatever data and techniques (e.g.
alternative interpolation techniques) are desired by any particular user.
5.1.1 Scenario Name
The Scenario Name allows you to uniquely identify each analysis that you conduct. It is
used in several ways.
Results file name. The results file is given the Scenario Name (e.g., Example Annual
PM.asr). Note that the extension f .ASR) is specifically designated just for MATS and
can only be used by MATS.
Organize output. In the Output folder, MATS will generate a folder using the Scenario
Name. MATS will use this folder as a default location for files generated with this
Scenario Name.
Abt Associates Inc.
82
June 2010
-------
Annual PM Analysis: Details
C:\Program Files\Abt Associates\MATS\output
ISM
File Edit View Favorites Tools Help
*
^Back ' Search
Folders
Em-
I Address £3 C:\Program Files\Abt Associates\MATS\output
V
|B»
Folders x
Name
Size Type
- l_j Program Files
Q Q Abt Assoc
E i£^) Abt Associates
B Q MATS
!h£) configs
data
. l£) help
maps
output
sarnpledata
r+l lr i i.MnrL;-
File Folder
Hi Example Annual PM.asr 169,020 KB ASR File
v
> <
Output file names. The output files generated will begin with the Scenario Name.
Help
Start Map View | Output Navigator I
Load
Extract All
Highlight file of interest and right-click to view options to Map, View, and Extract the data.
|Name
Type
| Size
t Confiauration/Loa Files
Configuration
Configuration
223kb
Log File
Run Log
1 kb
- Output Files
Example Annual PM Baseline Quarterly Avg Model Data
Monitor Network
2189kb
Example Annual PM Quarterly Avg Speciated Monitors
Monitor Network
140kb
Example Annual PM Quarterly Avg NH^/DON Monitors
Monitor Network
53kb
Example Annual PM Used Baseline Quarterly Avg Model Data Point
Monitor Network
604kb
Example Annual PM Quarterly Avg Spec Frac Point
Monitor Network
136kb
Example Annual PM Quarterly Avg Spec Frac Spatial Field
Monitor Network
3857kb
Example Annual PM Neighbor File Spatial Field
Monitor Network
138301 kb
Example Annual PM Grad Adj Quarterly Avg Spec Frac Spatial Field
Monitor Network
3853kb
Example Annual PM Future Quarterly Avg Model Data
Monitor Network
2182kb
Example Annual PM Used Future Quarterly Avg Model Data Point
Monitor Network
602kb
Example Annual PM Quarterly PM25 Point
Monitor Network
180kb
Example Annual PM Annual PM25 Point
Monitor Network
45kb
Example Annual PM High County Sites
Monitor Network
33kb
Example Annual PM Neighbor File Point
Monitor Network
4955kb
Example Annual PM Quarterly PM25 Spatial Field
Monitor Network
49E19kb
Example Annual PM Annual PM25 Spatial Field
Monitor Network
1248kb
Example Annual PM Grad Adj Quarterly PM25 Spatial Field
Monitor Network
4869kb
Example Annual PM Grad Adj Annual PM25 Spatial Field
Monitor Network
1246kb
stop info
Abt Associates Inc.
83
June 2010
-------
Annual PM Analysis: Details
5.1.2 Standard Analysis
Future-year PM2.5 design values are calculated at each FRM monitoring site through a
series of calculations:
Step 1. Baseline quarterly average PM2.5 calculation at each FRM monitor site;
Step 2. Baseline quarterly average species calculation using the quarterly weighted-
average baseline PM2.5 concentration and species fractions at each FRM monitor site;
Step 3. Forecasted quarterly average species calculation using relative response factors
(RRFs) at each FRM monitor site;
Step 4. Forecasted design value calculation at each FRM monitor site.
In this section, we go into some detail describing these steps. However, you should note
that MATS gives you a number of options affecting the exact steps that MATS follows,
such as determining which years of monitoring data to use and to choosing which monitors
to include in the calculations. These options are detailed in the PM2.5 Calculation Options
section. The output from the Standard Analysis is described here.
5.1.2.1 Step 1: Baseline Quartelry Average PM2.5 Calculation
The first step in the Standard Analysis is to calculate baseline PM2.5 levels using the
("official") quarterly average PM2.5 file (described in the Data Input section). MATS uses
these quarterly values to calculate 3-year averages of consecutive years of data for each
quarter, and then to average these averages to get a single PM2.5 estimate for each quarter.
Example Calculation Baseline PM2.5 Concentration
Starting with the following quarterly values:
Quarter
2000
2001
2002
2003
2004
Q1
11.8188
10.7107
9.7133
10.6966
9.9839
Q2
13.2400
11.1333
9.3138
15.1655
9.6680
Q3
18.7960
11.0700
12.6652
12.0667
14.4964
Q4
14.2579
9.4032
9.7903
10.7414
11.2778
MATS calculates 3-year averages of consecutive years of data for each quarter:
Quarter
2000-2002
2001-2003
2002-2004
Q1
10.7476
10.3735
10.1313
Q2
11.2290
11.8709
11.3824
Q3
14.1771
11.9340
13.0761
Q4
11.1505
9.9783
10.6032
MATS averages the 3-year averages to get a single estimate for each quarter:
Abt Associates Inc.
84
June 2010
-------
Annual PM Analysis: Details
Quarter
Avg
Q1
10.4175
Q2
11.4941
Q3
13.0624
Q4
10.5773
5.1.2.2 Step 2: Baseline Quarterly Average Species Calculation
Since the FRM monitors do not have speciated data (and the majority of FRM sites are not
co-located with a speciation monitor), MATS uses speciated PM2.5 monitor data from
other monitoring networks, such as STN and IMPROVE, to estimate the PM2.5
attributable to the following species: sulfate (S04), nitrate (N03), elemental carbon (EC),
organic carbon (OC), crustal matter, particle bound water (PBW), ammonium (NH4), and,
data permitting, salt.
Note that the process of calculating species fractions is involved and discussed in detail in
the Species Fractions Calculations & Output section of this user manual. Nevertheless, the
use of species fractions to calculate species concentrations is straightforward. The
weighted quarterly species average is calculated by multiplying weighted quarterly average
FRM baseline values (minus the assumed blank mass, specified by the user) with species
fractions that have been estimated for each FRM monitor. The calculation is as follows:
Species^ = Specie sFractioni£ ¦ i PM25g — BlankMass i
where.
Species^ = weighted quarterly average for a given species " i" (e. g. ,SOA)
Specie sFraction^ ¦ q = species fraction for species" z"
PM25£ = weighted quarterly average PM2 $
BlankMass = assumed monitoringblank mass{e. g. ,0.5ugf ml)
Note that MATS calculates species fractions from speciated monitors for a limited number
of years. As a result, rather than have species fractions specifically calculated for each
quarter and each year, MATS uses a single set of species fractions to calculate the
weighted quarterly average species concentrations. The species data should be
"representative" of the species fractions that occur during the 5 year FRM monitoring
period selected in MATS.
Example Calculation Baseline Species Concentrations
MATS multiplies the (non-blank) baseline quarterly PM2.5 values:
FRM PM2.5 Blank Mass Non-Blank Mass
10.4175 0.5 9.9175
11.4941 0.5 10.9941
Abt Associates Inc.
85
June 2010
-------
Annual PM Analysis: Details
13.0624 0.5 12.5624
10.5773 0.5 10.0773
with the species fractions calculated for this particular site:
Quarter
S04
N03
OCMMB
EC
PBW
NH4
Crustal
(Salt)
Q1
0.2382
0.1108
0.3346
0.0201
0.1049
0.1209
0.0705
--
Q2
0.2637
0.0727
0.3178
0.0545
0.1095
0.1273
0.0545
-
Q3
0.1432
0.0557
0.5621
0.0238
0.0959
0.0955
0.0238
-
Q4
0.2580
0.1389
0.2756
0.0396
0.1094
0.1190
0.0595
—
The product is the estimated baseline species concentrations by quarter.
Quarter
S04
N03
OCMMB
EC
PBW
NH4
Crustal
Q1
2.3623
1.0989
3.3184
0.1993
1.0403
1.1990
0.6992
Q2
2.8991
0.7993
3.4939
0.5992
1.2039
1.3996
0.5992
Q3
1.7989
0.6997
7.0613
0.2990
1.2047
1.1997
0.2990
Q4
2.5999
1.3997
2.7773
0.3991
1.1025
1.1992
0.5996
Baseline Calculation - General
MATS does several calculations to generate the species concentrations used to calculate
species fractions. In general, calculation of the concentrations of S04, N03, EC, crustal,
and salt are straightforward. The concentrations are derived directly from the ambient data
file or an interpolation of that data. However, the calculation of ammonium, particle
bound water (PBW) and organic carbon (OC) are more complicated and calculated
internally in MATS (as discussed in the following sections).
MATS uses speciated monitor data to estimate individual species fractions at FRM sites
using the "SANDWICH" process (Frank, 2006) (SANDWICH stands for Sulfates,
Adjusted Nitrates, Derived Water, Inferred Carbonaceous mass, and estimated aerosol
acidity [H+]). The data input to the PM calculations in MATS includes quarterly FRM
monitor data and speciated monitor data from STN and IMPROVE sites that has been
partially adjusted to match the anomalies in FRM data (e.g., nitrate volatilization).
The default species input data file contains aerosol nitrate data (N03r) that has been
adjusted to account for volatilization. Additional SANDWICH adjustments are made
within MATS. These include calculation of particle bound water (PBW) and organic
carbon by mass balance (OCMmb).
When there is more than one year of speciated data, MATS will create quarterly average
species levels for each year at each monitor, and then average the seasonal values across
the available years to get a single estimate for each species for each quarter at each
monitor. (See the section on Species Fractions Calculation Options for additional details
on how multiple years of speciated monitor data are combined.)
Abt Associates Inc.
86
June 2010
-------
Annual PM Analysis: Details
Note that MATS can calculate quarterly species concentrations at FRM monitor sites or a
spatial field. MATS will allow you to reuse the species fractions file for point estimates or
spatial fields. And in addition to calculating species fractions with monitor data — on the
Output Choice - Advanced window, you can choose gradient-adjusted species fractions,
which are based on monitor plus model data.
Interpolation
About 75 percent of FRM monitors are not co-located with an STN monitor, so the
estimation of the quarterly averages of the individual species at those FRM sites depends
on the interpolated quarterly averages from speciated monitors (e.g., STN). Individual
species are interpolated to the latitude and longitude associated with each FRM monitor.
(For FRM monitors that are co-located with an STN monitor, MATS simply uses the
species values from the co-located STN monitor.) You can find details on the
interpolation process and different options for interpolation at the section on Interpolation
Options for Species Fractions Calculation.
Calculations After Interpolation
After the interpolation of quarterly averages for retained N03 (N03r), S04, OCM, crustal,
EC, and DON, a few additional steps are necessary to generate speciated quarterly averages
at each FRM monitor site. These include calculating retained NH4 (NH4r), PBW, blank
mass, and organic carbon mass (OCMMB), the latter of which is calculated through a mass
balance approach.
Summary of Calculations After Interpolation of STN & IMPROVE Speciated Monitor Data
Calculation Description
Calculate Retained Ammonium Calculate ammonium associated with retained nitrate (N03r) and
(NH4r) S04. MATS calculates NH4r using DON, S04, and N03r.
(Alternatively MATS can use directed measured ammonium).
Calculate Particle Bound Water Calculate amount of water associated with ammonium sulfate and
(PBW) ammonium nitrate, which are hygroscopic.
Estimate Blank Mass Account for contamination in FRM monitors.
Calculate Organic Carbon Mass Calculate organic carbon mass with a mass balance approach.
(OCMMB)
Calculate Species Fractions Divide species estimates for S04, N03r, OCMMB, EC, crustal
material, NH4r, and PBW by the non-blank PM2.5 mass. (The
inclusion of salt is optional and is not included in the default MATS
data.)
The following sections describe the calculations of ammonium, PBW, and organic carbon
(by difference).
Retained Ammonium Calculation
MATS calculates retained ammonium two different ways. The default approach is to use
interpolated degree of neutralization of sulfate (DON) values from the speciated monitors.
The alternative approach is to use interpolated NH4 values from speciated monitors (e.g.,
Abt Associates Inc.
87
June 2010
-------
Annual PM Analysis: Details
STN). In the Species Fractions Calculations Options - Advanced section, you have the
option to choose the approach that you prefer to use. The two approaches are described
here.
Default approach using measured pre-calculated DON, S04, and retained N03
(N03r):
Because of uncertainties in NH4 speciation measurements, by default MATS calculates
ammonium values using the degree of sulfate neutralization (DON). MATS uses pre-
calculated daily DON values that are included in the species data input file ("Species-for-
fractions-xxxx.csv"). The values for DON are calculated from the amount of ammonium
associated with sulfate (NH4SQ4) as follows:
DON-
dun - ySQ
And the estimated NH4SQ4 is calculated as follows:
= NH^MeajUKd - 0.29 * M}^reaixd
where 0.29 is the molar ratio of NH4 to N03 and NH4measured and N03retained reflect the
amounts of NH4 and N03 retained on the FRM filter. The amount of NH4SQ4 is not
allowed to exceed the fully neutralized amount of 0.375 multiplied by the estimated sulfate
ion concentration.
MATS then calculates ammonium using interpolated monitor values of DON, S04, and
N03r as follows:
NH, = DON* SO, + 0.29*
Alternative Approach Using Measured Ammonium. The alternative approach is to use
interpolated NH4 values from STN monitors. This approach has several steps.
First, MATS calculates "adjusted" NH4:
~ {PctEvap *0.29 *{NOlMeasirsd - N03^d)
where the PctEvap factor refers to the percentage of ammonium associated with the
volatilized nitrate that is also lost. (As discussed in the Species Fractions Calculation
Options - Advanced section, this factor is adjustable from 0 to 100 percent.) The default
assumption is that no ammonium is volatilized (0 percent).
Second, MATS calculates NH4 associated with S04:
NH,iSn = -0.29*^0,^
Third, MATS calculates DON:
Abt Associates Inc.
88
June 2010
-------
Annual PM Analysis: Details
DON _NIi*,so*/
UUNcaktimed ~ /S0A
Finally, using the same equation as in the default approach, MATS calculates NH4r by
substituting the calculated DON for the interpolated (measured) DON value:
= ^dcylamd *^4 + 0.29*
Particle Bound Water Calculation
Because ammoniated sulfate and ammonium nitrate are hygroscopic, the retained sulfate
and nitrate mass will include water. Particle bound water (PBW) is estimated using the
Aerosol Inorganic Model (AIM) (Clegg et al, 1998). For computational convenience, a
polynomial regression equation was fit to the calculated water mass from AIM and the
three input values that fed into AIM (sulfate, nitrate and ammonium). AIM was run with
typical FRM filter equilibration conditions of 35% RH and 22 deg C (295 deg K).
MATS calculates particle-bound water (PBW) using levels of S04, N03r, and NH4r as
follows. (Note that this is the same equation that MATS uses to calculate future-year
PBW, the difference being the future-year PBW uses future-year values of S04, N03r, and
NH4r, and here MATS uses base-year values.)
The calculation uses one of two equations, depending on the acidity of the ammoniated
sulfate (represented by DON). S, N, and A in the equations are the relative fraction of
S04, N03r, and NH4r respectively.
S = S04 / (S04 + N03r + NH4r);
N = N03r / (S04 + N03r + NH4r);
A = NH4r / (S04 + N03r + NH4r);
ifDONle 0.225 then
PBW = {595.556
- 1440.584*S
- 1126.488*N
+ 283.907*(S**1.5)
- 13.384*(N**1.5)
- 1486.711*(A**1.5)
+ 764.229*(S**2)
+ 1501.999*(N*S)
+ 451.873*(N**2)
Abt Associates Inc.
89
June 2010
-------
Annual PM Analysis: Details
- 185.183*(S**2.5)
-375.984*(S**1.5)*N
- 16.895*(S**3)
- 65.814*(N**1.5)*S
+ 96.825*(N**2.5)
+ 83.037*(N**1.5)*(S**1.5)
- 4.419*(N**3)
+ 1720.818*(A** 1.5)* S
+ 1220.383*(A**1.5)*N
- 311.496*(A**1.5)*(S**1.5)
+ 148.771*(A**1.5)*(N**1.5)
+ 1151.648*(A**3)} * (S04+N03r+NH4);
ELSE
PBW = {202048.975
-391494.647 *S
-390912.147 *N
+ 442.435 *(S**1.5)
- 155.335 *(N**1.5)
-293406.827 *(A**1.5)
+ 189277.519 *(S**2)
+ 377992.610 *N*S
+ 188636.790 *(N**2)
-447.123 *(S**2.5)
- 507.157 *(S**1.5)*N
- 12.794 *(S**3)
+ 146.221 *(N**1.5)*S
+ 217.197 *(N**2.5)
+ 29.981 *(N**1.5)*(S**1.5)
Abt Associates Inc.
90
June 2010
-------
Annual PM Analysis: Details
- 18.649 *(N**3)
+ 216266.951 *(A**1.5)*S
+ 215419.876 *(A**1.5)*N
-621.843 *(A**1.5)*(S**1.5)
+ 239.132 *(A**1.5)*(N**1.5)
+ 95413.122 *(A**3)} * (S04+N03r+NH4) .
Organic Carbon Mass Calculation
Measured organic carbon mass is not directly used in the calculation of species fractions in
MATS because of (1) many uncertainties in estimating carbonaceous mass from carbon
measurements (Turpin & Lim, 2001; Chow et al, 2004) (2) differences in carbon
measurement protocol between urban and rural monitoring locations, (3) a relatively
"bumpy" surface of urban carbon concentrations as derived from urban and rural organic
carbon measurements and (4) lack of carbon measurements at all FRM locations. The
MATS approach estimates carbon by mass balance comparing precisely measured FRM
PM2.5 mass (EPA, 2003) with the sum of its non-carbon components.
Total carbonaceous mass contains both elemental carbon (EC) and organic carbon mass
(OCM). We measure EC from the interpolated STN and IMPROVE monitors, while we
calculate OCM using a mass balance approach — and refer to it as OCMMB. To calculate
OCMMB, we subtract the other estimated retained species (including EC) from the PM2.5
level measured at the FRM site as follows:
OCM&m = PM25 - j SO^ + NOx + returned + PBW + Crustal + EC + Blank Mass+ (Salt)}
The value for OCMMB could be very small, or even be calculated as negative (if the sum
of the species enclosed in the curly brackets exceeded the FRM PM2.5 monitor value). To
ensure that the OCMMB does not get too small, an OCMMB "mass balance
floor" (default) value is set to 1.0 times the interpolated value of blank-adjusted organic
carbon (OCb). It is also possible that the value of the floor by itself could exceed the FRM
total PM2.5 value. In this case, MATS imposes a (user-adjustable) "ceiling," such that
OCMMB does not exceed a percentage of the total non-blank mass. The default ceiling
value is set to 0.8 or 80% of PM2.5 mass. (You can modify the floor and ceiling
assumptions in the Species Fractions Calculation Options - Advanced window.)
To account for these possibilities, MATS uses the following series of equations to
calculate OCMMB:
Abt Associates Inc.
91
June 2010
-------
Annual PM Analysis: Details
0CMmb,juiia; = NonBlankMass- ¦] +NOl retBixed +NHi KlBixed + PBW + Crustal + EC + Sah\
OCM^mer = Floor* OCb
(JCtivmrtiax = Max{0Fber'O^MB, Jisiai)
OCM!tS£ aai = Ceiling*NonBlankMass
OCM^
,J5 BOi
M«( OCM ^ Cci&xg'O
where the "Floor" variable has a default value in MATS of "1.0" and the "Ceiling" variable
has a default value in MATS of "0.8".
There are at least two things to note with this approach. (1) When the final OCMMB
value is equal to the "floor," then the sum of the species will exceed the PM2.5 value at the
FRM monitor. To ensure that the sum of the species just equals the FRM PM2.5 value,
MATS reduces all of the species (except OCMMB) by the same percentage until the sum
of the species just equals the FRM PM2.5 value. (2) When the final OCMMB value is
equal to the "ceiling," then the sum of the species will be less than the PM2.5 value at the
FRM monitor. In that case, MATS increases all of the species by the same percentage
until the sum of the species just equals the FRM PM2.5 value
Blank Mass Assumption
The field blank typically has a value of between 0.3 and 0.5 ug/m3, which appears to result
from contamination of the FRM filter. For calculating retained PM2.5, MATS uses a
default blank mass value of 0.5 ug/m3. If desired, you can change the default blank mass
value at the Species Fractions Calculations Options - Advanced window.
5.1.2.3 Step 3: Forecasted Quarterly Average Species Calculation
To calculate forecasted quarterly species averages for each of the four quarters in a given
year, MATS uses weighted average quarterly species concentrations and both baseline and
forecasted (e.g., 2020) air quality modeling. Because this calculation involves modeling
data from two different years, this is referred to as a "temporal adjustment."
The forecasted weighted quarterly average for each species is calculated by multiplying the
baseline weighted quarterly average for each monitor species with the ratio of the modeling
data. This process gives forecasted weighted quarterly averages of six species: S04, N03,
OCM, EC, crustal material, and NH4. The form of the equation is as follows:
Abt Associates Inc.
92
June 2010
-------
Annual PM Analysis: Details
Mo del ¦ fmre0
= ^>eciesi/auQ ¦
Model ijfoxQ
where
Speciesijl^SVQ - Estimated forecasted species"average in quarter Q
Speciesi^as£Q = Monitored baseline species " i" average in quarter Q
Model ijltur^Q = Modeled forecasted species "i" average in quarter Q
Model= Modeled baseline species " i" average in quarter Q
In other words, baseline species concentrations are assumed to change in the same
proportion as the model data in the same location from the baseline to the forecasted.
These proportions, called relative response factors (RRF), are simply the ratio of the
modeled forecasted species to the modeled baseline species.
RRF _ Model^Q
UB Models
where
RRFq = relative response factor for quarterQ
MATS calculates RRFs for each quarter for each of six species: S04, N03, OCM, EC,
crustal material, and (optionally) NH4. (Additional information on the calculation of the
RRFs can be found here.) The calculation of forecasted weighted quarterly average species
concentrations can be rewritten as:
Specie St = Species^^ ¦ RRFig
Additional calculations are needed to estimate future-year quarterly averages of NH4 and
particle-bound water (PBW), which is calculated using forecasted levels of NH4, N03, and
S04. (Details on the PBW calculation can be found here.)
Recall that the default base year NH4 calculation is as follows:
NH, = DON* SO, + 0.29* NOMed
MATS can calculate the future year NH4 concentration using modeled NH4 RRFs, or by
using the the base year DON value combined with future year S04 and N03 values
(default approach) as follows:
NH 4 = DON lose * SO 4, future + 0.29* NO-^ ^tllK
The option for choosing which approach to use for calculating future NH4 is given in the
PM2.5 Calculation Options window. Finally, note that PBW is calculated after future year
NH4, using the previously identified water equation and future year concentrations of
NH4, S04, and N03.
Abt Associates Inc.
93
June 2010
-------
Annual PM Analysis: Details
Example Calculation Forecasted Species Concentrations
MATS multiplies the baseline species concentrations:
Quarter
Q1
Q2
Q3
Q4
S04
2.3623
2.8991
1.7989
2.5999
N03
1.0989
0.7993
0.6997
1.3997
OCMMB
3.3184
3.4939
7.0613
2.7773
EC
0.1993
0.5992
0.2990
0.3991
PBW
1.0403
1.2039
1.2047
1.1025
NH4
1.1990
1.3996
1.1997
1.1992
Crustal
0.6992
0.5992
0.2990
0.5996
with the RRFs for each quarter and species:
Quarter
S04
N03
OCMMB
EC
PBW
NH4
Crustal
Q1
0.9737
0.9873
0.9636
0.9872
-
0.9634
0.9808
Q2
0.9898
0.9991
0.9979
0.9917
-
0.9697
0.9891
Q3
0.9784
0.9775
0.9875
0.9944
-
0.9702
0.9759
Q4
0.9700
0.9800
0.9761
0.9843
—
0.9790
0.9818
The product is the forecasted species concentrations by quarter.
Quarter
S04
N03
OCMMB
EC
PBW
NH4
Crustal
Q1
2.3003
1.0849
3.1975
0.1968
-
1.1552
0.6858
Q2
2.8695
0.7985
3.4865
0.5942
-
1.3571
0.5927
Q3
1.7600
0.6840
6.9728
0.2973
-
1.1640
0.2918
Q4
2.5219
1.3717
2.7109
0.3928
—
1.1741
0.5887
5.1.2.4 Step 4: Forecasted Design Value Calculation
The forecasted weighted quarterly averages for the species are then added together to get
the estimated forecasted quarterly average PM2.5 level:
PM25,q = EC\,q + M03,q + OCMg + BCq + Crustalg + NH^Q + PB WQ
MATS averages the four quarterly averages these forecasted quarterly average PM2.5
levels to calculate forecasted design values:
4 PM23D
pm2S = y —^
8
where
PM23 - annual awrage PM25 design value.
The output file containing the results of this Standard Analysis are described here.
Example Calculation Forecasted Design Values
Abt Associates Inc.
94
June 2010
-------
Annual PM Analysis: Details
Using hypothetical data from the previous example calculation plus PBW values and
assumed blank mass, MATS sums the species to get quarterly PM2.5 values.
Quarter S04 N03 OCMMB EC PBW NH4 Crustal Blank PM2.5
Mass
Q1 2.3003 1.0849 3.1975 0.1968 0.5000 1.1552 0.6858 0.5000 9.6205
Q2 2.8695 0.7985 3.4865 0.5942 0.4000 1.3571 0.5927 0.5000 10.5985
Q3 1.7600 0.6840 6.9728 0.2973 0.6000 1.1640 0.2918 0.5000 12.2699
Q4 2.5219 1.3717 2.7109 0.3928 0.7000 1.1741 0.5887 0.5000 9.9601
MATS then averages the four quarters to get a forecasted design value:
Quarter PM2.5
Q1 9.6205
Q2 10.5985
Q3 12.2699
Q4 9.9601
Forecasted 10.61
Design Value
5.1.2.5 Output Description
The output file is named "AnnualPM25 Point.csv" with the Scenario Name appended at
the beginning and the forecast year is inserted at the end (e.g., "Example PM — Annual
PM25 Point 2020.csv"). The table below describes the variables in the output file for the
Standard Analysis.
Note: The RRF variables in this file are not the actual RRFs used to calculate future year
PM2.5 and PM2.5 species. They are the resultant annual average RRFs calculated by
dividing the future annual average concentrations (in this file) by the base year annual
average concentrations (in this file). The actual RRFs are calculated on a quarterly
average basis and are contained in the quarterly average output files. There are no
quarterly average RRFs for water or NH4 (if DON, N03, and S04 are used to calculate
NH4).
Output file name: "Scenario Name + Annual PM25 Point"
Variable Description
Jd The ID is a unique name for each monitor in a particular location. The default
value is the AIRS ID. (This is a character variable.)
_type FRM data
_STATE_NAME State name. (This is a character variable.)
_COUNTY_NAM County name. (This is a character variable.)
E
monitorjat Latitude at the monitor site in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are negative.
Abt Associates Inc.
95
June 2010
-------
Annual PM Analysis: Details
monitorjong Longitude at the monitor site in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g., United
States) are negative.
monitor_gridcell Identifier of grid cell closest to the monitor
b_pm25_ann_DVBase year 5 year weighted average PM2.5 annual design value
f_pm25_ann_DV Future year 5 year weighted average PM2.5 annual design value
b_blank_mass Base year blank mass concentration (ug/m3)
b_crustal_mass Base year crustal mass concentration (ug/m3)
b_EC_mass Base year elemental carbon mass concentration (ug/m3)
b_NH4_mass Base year ammonium mass concentration (ug/m3)
b_Ocmb_mass Base year organic carbon mass (by difference) concentration (ug/m3)
b_S04_mass Base year sulfate ion mass concentration (ug/m3)
b_N03_mass Base year nitrate ion mass concentration (ug/m3)
b_water_mass Base year water mass concentration (ug/m3)
b_salt_mass Base year salt mass concentration (ug/m3)
f_blank_mass Future year blank mass concentration (ug/m3)
f_crustal_mass Future year crustal mass concentration (ug/m3)
f_EC_mass Future year elemental carbon mass concentration (ug/m3)
f_NH4_mass Future year ammonium mass concentration (ug/m3)
f_Ocmb_mass Future year organic carbon mass (by difference) concentration (ug/m3)
f_S04_mass Future year sulfate ion mass concentration (ug/m3)
f_N03_mass Future year nitrate ion mass concentration (ug/m3)
f_water_mass Future year water mass concentration (ug/m3)
f_salt_mass Future year salt mass concentration (ug/m3)
rrf_crustal Resultant annual relative response factor- Crustal Mass
rrf_ec Resultant annual response factor- Elemental Carbon Mass
rrf_nh4 Resultant annual relative response factor- Ammonium Mass.
rrf_oc Resultant annual relative response factor- Organic Carbon Mass
rrf_so4 Resultant annual relative response factor- Sulfate Mass
rrf_no3 Resultant annual relative response factor- Nitrate Mass
rrf_water_mass Resultant annual relative response factor- Water Mass
rrf salt
Resultant annual relative response factor- Salt Mass (set equal to 1 if modeled
salt is not used)
5.1.3 Quarterly Model Data
The Quarterly Model Data option gives you the option of creating a small, reusable file
with quarterly values from a much larger file with daily values. Since annual PM2.5
MATS works with quarterly values there is no loss of precision. To save time, it is
Abt Associates Inc.
96
June 2010
-------
Annual PM Analysis: Details
possible to use daily values only for an initial run with MATS and check this Quarterly
Model Data option. Then for subsequent runs (that use the same modeled data), use the
quarterly file that MATS generates. However, this will only work for subsequent MATS
runs that use exactly the same base and future year photochemical model data (such as
sensitivity runs that test the various ambient data settings in MATS).
Alternatively, you can generate a baseline and future quarterly average model file outside
of MATS (using a program such as SAS, STAT A, etc), and then load these quarterly files
into MATS, bypassing the use of any daily model data in MATS. The format of the
quarterly file is described below.
MATS generates two files:
Base-year file: "Baseline Quarterly AvgModelData.csv" with the Scenario Name (e.g.,
Example PM Annual) appended at the beginning (e.g., "Example PM Annual — Baseline
Quarterly Avg Model Data.csv").
Future year file (e.g., "Example PM — Future Quarterly Avg Model Data. csv." The
format of the two files is the same.
The table below describes the variables in the Quarterly Model Data file.
Output file name: "Scenario Name + Baseline/Future Quarterly Avg Model Data"
Variable Description
Jd The ID is a unique identifier for each model grid cell. The default value is the
column identifier multiplied by 1000 plus the row. (This is a character variable.)
Jype
lat Latitude at the grid cell centroid in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are negative.
long Longitude at the grid cell centroid in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g., United
States) are negative,
date Year and Quarter (01= 1st quarter, 04= 2nd quarter, etc.)
crustal Crustal PM
nh4 Ammonium PM
so4 Sulfate PM
ec Elemental Carbon
no3 Nitrate PM
oc Organic carbon PM
pm25 PM2.5 mass (only used to gradient adjust PM2.5 for gradient adjusted spatial
fields)
cm Coarse PM (ug/m3) (only used for visbility calculations)
MATS also produces a set of "used" quarterly average model day files (base and future).
Abt Associates Inc.
97
June 2010
-------
Annual PM Analysis: Details
These have the same format as the quarterly average model data files, but only contain data
for the model grid cells that were used in the MATS point calculations. These files can
also be re-used and are also useful for QA purposes.
The table below describes the variables in the Used Quarterly Model Data file:
Output file name: "Scenario Name + Used Baseline/Future Quarterly Avg Model
Data"
Variable Description
Jd The ID is a unique identifier for each model grid cell. The default value is the
column identifier multiplied by 1000 plus the row. (This is a character variable.)
Jype
lat Latitude at the grid cell centroid in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are negative.
long Longitude at the grid cell centroid in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g., United
States) are negative,
date Year and Quarter (01= 1st quarter, 04= 2nd quarter, etc.)
crustal Crustal PM
nh4 Ammonium PM
so4 Sulfate PM
ec Elemental Carbon
no3 Nitrate PM
oc Organic carbon PM
pm25 PM2.5 mass (only used to gradient adjust PM2.5 for gradient adjusted spatial
fields)
cm Coarse PM (ug/m3) (only used for visbility calculations)
5.1.4 Species Fractions
Species fractions are simply the fraction of quarterly average PM2.5 at a given monitor
attributable to seven (and potentially eight) species: nitrate (N03), sulfate (S04), organic
carbon (OC), crustal, elemental carbon (EC), ammonium (NH4), and particle-bound water
(PBW). (And pending data availability, an eighth species, salt, can be included as well, the
default MATS species files include salt data. But salt is an optional species on the model
files. If base and future year modeled salt data is supplied, a salt RRF will be calculated.
If there is no salt data in the model files, then the salt RRF will be set to 1.)
When you check the Species Fractions option, you get a reusable file with species fractions
for each FRM monitor. Making the file reusable allows you to generate consistent results
and, perhaps most importantly, allows the same file to be used by different MATS users.
5.1.4.1 Species Fractions Calculation
After calculating the ambient level of S04, N03, OCMMB, EC, PBW, NH4, and crustal,
Abt Associates Inc.
98
June 2010
-------
Annual PM Analysis: Details
MATS then divides these ambient levels by the non-blank mass of PM2.5. To get non-
blank PM2.5, MATS subtracts the blank mass from the FRM PM2.5 value. MATS then
divides each of the species (except blank mass) by non-blank PM2.5.
Example Calculation of Species Fractions
The table below gives an example of theses calculations. The fraction is calculated by
dividing mass (ug/m3) by the Non-blank Mass. Note that salt is optional.
Units FRM Blank Non-BIa S04 N03 OCMM EC PBW NH4 Crusta (Salt)
PM2.5 Mass nk Mass B I
Concentrati 10.9175 0.5 10.4175 2.41601.1555 3.4847 0.2101 1.15551.26050.7353 -
on
Fraction 0.23190.1109 0.3345 0.0202 0.1109 0.1210 0.0706 -
5.1.4.2 Output Description
The output file is named "Quarterly Avg Spec Frac Point.csv" with the Scenario Name
appended at the beginning (e.g., "Example PM — Quarterly Avg Spec Frac Point.csv").
The table below describes the variables in the file.
The interpolated variables (starting with i xxx) are created when MATS is run, but are not
needed when re-using a fractions file. They are also not needed when running with a user
generated fractions file.
Output file name: "Scenario Name + Quarterly Avg Spec Frac Point"
Variable
_id
_STATE_NAME
_COUNTY_NAME
monitorjat
monitorjong
quarter
pm25_mass_frac
fcr
fee
fnh4
focm
fso4
fno3
fwater
Description
The ID is a unique name for each monitor in a particular location.
The default value is the AIRS ID. (This is a character variable.)
State name. (This is a character variable.)
County name. (This is a character variable.)
Latitude at the monitor site in decimal degrees. Values in the
northern hemisphere are positive, and those in the southern
hemisphere are negative.
Longitude at the monitor site in decimal degrees. Values in the
eastern hemisphere are positive, and those in the western
hemisphere (e.g., United States) are negative.
Quarter
PM2.5 mass used to calculate species fractions (calculated from
the "PM2.5 for fractions" file)
Crustal fraction of PM2.5 mass
Elemental carbon fraction of PM2.5 mass
Ammonium fraction of PM2.5 mass
Organic carbon fraction of PM2.5 mass
Sulfate ion fraction of PM2.5 mass
Nitrate ion fraction of PM2.5 mass
Water fraction of PM2.5 mass
Abt Associates Inc.
99
June 2010
-------
Annual PM Analysis: Details
fsalt
Salt fraction of PM2.5 mass
Blank mass
Degree of neutralization of sulfate used to calculate ammonium
mass (0.000 - 0.375)
Interpolated sulfate ion
Interpolated retained nitrate ion
Interpolated blank-adjusted organic carbon
Interpolated elemental carbon
Interpolated crustal
Interpolated degree of neutralization of sulfate (DON).
Interpolated ammonium
Interpolated nitrate ion (only used to calculate volatilized
ammonium; if option is selected)
Interpolated salt
blank mass
don
so4
no3r
ocb
ec
crustal
don
nh4
no3
i salt
Notes:
i ocb is only used to to calculate the OCMmb "floor".
i_nh4 is not used if DON is used to calculate the ammonium concentration (and fraction).
i_no3 is only used to calculate the "volatilized ammonium", if the option is selected (not
used by default).
In the Output Choice Advanced window, MATS lets you choose from among two main
options: Spatial Field Estimates and Miscellaneous Output that is generally used for
quality assurance (QA). Within each of these two main options there are a number of
choices.
5.2 Output Choice - Advanced
Abt Associates Inc.
June 2010
100
-------
Annual PM Analysis: Details
Annual PM Analysis
EH51 Choose Desired Output
Output Choices - Advanced
¦ Data Input
¦ Species Fractions Options
¦ Species Fractions - Advanced
¦ PM2.5 Calculation Options
¦ Model Data Options
Final Output and Check
Output Choice - Advanced
Spatial Field Estimates
Forecast
|7 Interpolate FRM and speciation monitor data to spatial field. Temporally adjust.
|7 Interpolate gradient-adjusted FRM and speciation monitor data to spatial field. Temporally adjust.
Miscellaneous Outputs
Quarterly average files
|7 Point [7 Spatial Field
[7 Spatial Field - gradient-adjusted
High county sites
|7 File "C"
Species fractions spatial field
|7 Spatial Field |7 Spatial Field - gradient-adjusted
Quarterly average speciated monitors
|7 File "E"
Neighborfiles
|7 Point [7 jSpatiai Field
Spatial Field Estimates This option gives you PM2.5 forecasts for each grid cell in
the modeling domain (e.g., CMAQ 36 km).
The Interpolate FRM & speciation monitor data to spatial field. Temporallv-adiust
option calculates interpolated spatial fields that are temporally adjusted. This option
creates gridded spatial fields of future year PM2.5 data. To create PM2.5 spatial fields,
MATS interpolates both speciation data and PM2.5 data (FRM and IMPROVE).
The Interpolate gradient-adjusted FRM & speciation monitor data to spatial field.
Temporallv-adiust option calculates interpolated spatial fields that are temporally
adjusted and gradient adjusted. Check this option to create gridded spatial fields of
gradient adjusted future year PM2.5 data. To create PM2.5 spatial fields, MATS
interpolates both gradient adjusted speciation data and PM2.5 data (FRM and
IMPROVE). This option creates the recommended spatial field for the "Unmonitored
Area Analysis" from the PM2.5 modeling guidance.
Miscellaneous Output. This provides a variety of output files that can be used to QA
the your calculations.
Quarterly average files. The MATS default is to output the annual average results for
all analyses (point, spatial fields). Checking any one of these boxes will also output the
more detailed quarterly average calculations (the quarterly average calculations are the
basis of all of the MATS annual PM2.5 calculations).
Abt Associates Inc.
101
June 2010
-------
Annual PM Analysis: Details
High county sites. The MATS default is to output the point results for all FRM sites.
Checking this box will also create a file which contains only the single highest monitor
in each County (based on the highest future year value). This dataset is a subset of the
all sites file.
Species fractions spatial field. This option is the same as the species fraction file
created from the standard analysis except it outputs the species fractions file created
from a spatial field or gradient adjusted spatial field. The file will contain species
fraction data for each quarter for each grid cell.
Quarterly average speciated monitors. This file contains the raw quarterly average
speciated data that MATS uses to do interpolations (to calculate species fractions). This
data is derived from the "species for fractions" input file.
Neighbor files. The neighbor files contain the "nearest neighbor" data for the VNA
interpolation scheme. The data includes the distance to neighbor monitors and weights
used to do the interpolations. There is information for each FRM monitor (for point
analyses) or each grid cell (for spatial fields) for each quarter and for each species.
5.2.1 Spatial Field Estimates
With the Spatial Field Estimates option you can estimate PM2.5 for each grid cell in a
spatial field by interpolating both FRM monitor data and speciated monitor data to the
grid cell centroid (identified by latitude and longitude). A key difference between Spatial
Field Estimates and Standard Analysis is that the Spatial Field Estimates requires the
interpolation of FRM monitor data. MATS interpolates the "unofficial" daily PM2.5
monitor data from FRM and IMPROVE sites to each grid cell centroid and then averages
the values for each quarter from the different PM2.5 sites. (Recall, as discussed in the
Baseline PM2.5 Calculation section, the Standard Analysis uses "official" PM2.5 monitor
data.)
The basic form of the spatial calculation is as follows:
Gride ell s toigss? = y Weighti ¦ Monitor
2-1
Gridcell, baseline = estimated quarterly average baseline species concentration at grid cell E;
Weight; = inverse distance weight for monitor i;
Monitor; = quarterly average baseline species concentration at monitor i.
Note that MATS lets you choose whether to use air quality modeling to scale the
interpolation of speciated monitor values to grid cell centroids (spatial fields). This use of
modeling data is referred to as a gradient adjustment, and is discussed in the next section.
Abt Associates Inc.
102
June 2010
-------
Annual PM Analysis: Details
5.2.1.1 Gradient-Adjustment- ("Fused fields")
Using modeling data for a gradient adjustment is fairly simple. MATS calculates species
fractions using monitor data in basically the same way as in the species fractions
calculation for the Standard Analysis, however, MATS includes model data to help inform
the calculation. In particular, MATS uses model values for the grid cell of interest and the
model values for the grid cells containing the speciated/PM2.5 monitors to be interpolated
to cell of interest. (Note that "unofficial" PM2.5 values are used in the interpolation, and
they are treated in essentially the same way as monitor data for particular species, such as
S04.)
A general form of the equation for the interpolated species values is as follows:
Specie sE = y Weigk^ ¦ Mom to¦ Gradient Adjustment B
i-l
where:
SpeciesE baseline = estimated baseline quarterly average species/PM2.5
concentration at cell E;
Weight; = inverse distance weight for monitor i;
Monitor; = baseline quarterly average species/PM2.5 concentration at
monitor i;
Gradient Adjustment; E = gradient adjustment from monitor i to cell E.
This interpolation calculation can be rewritten somewhat more precisely as:
c, ¦ "v-i ij57 ¦ •,
Specie sg = X Weight ' Momtor% ¦
i-l Moaek^e
where:
Model^ baseline = modeled quarterly average baseline species/PM2.5 concentration at cell E;
Model; baseline = modeled quarterly average baseline species/PM2.5 concentration at
monitor site i.
MATS uses this gradient approach for PM2.5 and for most species — S04, N03r, EC,
OCb, Crustal, and NH4. However, for DON (degree of neutralization), MATS does not
use a gradient adjustment and simply interpolates the estimated DON values as is from
each monitor. Options regarding the use of model data are discussed in more detail in the
section on Model Data Options.
Once MATS has interpolated quarterly species values at the cell of interest, calculations
Abt Associates Inc.
103
June 2010
-------
Annual PM Analysis: Details
proceed in exactly the same way as described for species fractions and forecasted PM2.5
values for the Standard Analysis.
Example Calculation for Interpolated Gradient-Adjusted Baseline Species
Concentrations
For each quarter, MATS identifies the speciated monitors that are nearby ("neighbors") to a
given cell of interest "E", and determines the weight that should be given to each monitor
in the interpolation process. Monitors that are nearby get a greater weight.
Speciated Monitor Values
Monitor Quarter S04 N03r EC OCb Crustal NH4 DON Interpolation
ID Wght
A Q1 2.30 1.10 3.32 0.20 1.10 1.20 0.21 0.2
B Q1 2.90 0.80 3.49 0.60 1.20 1.40 0.19 0.4
C Q1 1.80 0.70 7.06 0.30 1.20 1.20 0.18 0.15
D Q1 2.60 1.40 2.78 0.40 1.10 1.20 0.32 0.25
MATS then identifies the speciated model values at each of the speciated monitors (A-D)
and for the given cell of interest (E):
Model Values at Speciated Monitor Sites A-D
Monitor ID Quarter S04 N03 EC OC Crustal NH4 DON
A Q1 1.60 1.03 2.99 0.15 0.75 0.86
B Q1 2.54 0.70 3.20 0.47 1.07 1.04
C Q1 1.66 0.59 5.91 0.23 1.13 0.81
D Q1 1.93 1.15 2.13 0.27 1.00 1.12
Model Values at Cell E
Monitor ID Quarter S04 N03 EC OC Crustal NH4 DON
E Q1 2.10 0.98 3.10 0.15 1.00 1.30
MATS then calculates the speciated values at cell E using the monitor and model data (and
interpolation weights) using the equation above:
Interpolated Speciated Monitor Values
Monitor ID
Quarter
S04
N03r
EC
OCb
Crustal
NH4
DON
A
Q1
3.02
1.04
3.44
0.20
1.46
1.82
0.21
B
Q1
2.40
1.12
3.38
0.19
1.13
1.75
0.19
C
Q1
2.27
1.16
3.71
0.19
1.06
1.93
0.18
D
Q1
2.83
1.19
4.05
0.22
1.10
1.39
0.32
Wght. Avg.
Q1
2.61
1.13
3.61
0.20
1.18
1.70
0.23
Note that MATS does a similar calculation (not shown) when interpolating "unofficial"
PM2.5 values to the given cell of interest, denoted "E".
Abt Associates Inc.
104
June 2010
-------
Annual PM Analysis: Details
5.2.1.2 Output Description - Interpolate FRM & Speciation Monitor Data to Spatial
Field
The output file is named "Annual PM25 Spatial Field' with the Scenario Name appended
at the beginning and the forecast year is inserted at the end (e.g., "Example PM — Annual
PM25 Spatial Field 2020. csv"). The table below describes the variables in the output file.
Output file name: "Scenario Name + Annual PM25 Spatial Field"
Variable
Jd
gridcelljat
gridcelljong
b_pm25_ann_DV
f_pm25_ann_DV
b_blank_mass
b_crustal_mass
b_EC_mass
b_NH4_mass
b_Ocmb_mass
b_S04_mass
b_N03_mass
b_water_mass
b_salt_mass
f_blank_mass
f_crustal_mass
f_EC_mass
f_NH4_mass
f_Ocmb_mass
f_S04_mass
f_N03_mass
f_water_mass
f_salt_mass
rrf_crustal
rrf_ec
rrf_nh4
rrf_oc
rrf so4
Description
The ID is a unique identifier for each model grid cell. The default value is the
column identifier multiplied by 1000 plus the row. (This is a character
variable.)
Latitude at the grid cell centroid in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are negative.
Longitude at the grid cell centroid in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g., United
States) are negative.
Base year interpolated PM2.5 from unofficial data file
Future year interpolated PM2.5
Base year blank mass concentration (ug/m3)
Base year crustal mass concentration (ug/m3)
Base year elemental carbon mass concentration (ug/m3)
Base year ammonium mass concentration (ug/m3)
Base year organic carbon mass (by difference) concentration (ug/m3)
Base year sulfate ion mass concentration (ug/m3)
Base year nitrate ion mass concentration (ug/m3)
Base year water mass concentration (ug/m3)
Base year salt mass concentration (ug/m3)
Future year blank mass concentration (ug/m3)
Future year crustal mass concentration (ug/m3)
Future year elemental carbon mass concentration (ug/m3)
Future year ammonium mass concentration (ug/m3)
Future year organic carbon mass (by difference) concentration (ug/m3)
Future year sulfate ion mass concentration (ug/m3)
Future year nitrate ion mass concentration (ug/m3)
Future year water mass concentration (ug/m3)
Future year salt mass concentration (ug/m3)
Resultant annual relative response factor- Crustal Mass
Resultant annual response factor- Elemental Carbon Mass
Resultant annual relative response factor- Ammonium Mass.
Resultant annual relative response factor- Organic Carbon Mass
Resultant annual relative response factor- Sulfate Mass
Abt Associates Inc.
105
June 2010
-------
Annual PM Analysis: Details
rrf_no3 Resultant annual relative response factor- Nitrate Mass
rrf_water_mass Resultant annual relative response factor- Water Mass
rrf_salt Resultant annual relative response factor- Salt Mass
5.2.1.3 Output Description - Interpolate Gradient-Adjusted FRM & Speciation
Monitor Data to Spatial Field
The output file is named "Grad Adj Annual PM25 Spatial Field' with the Scenario Name
appended at the beginning and the forecast year is inserted at the end (e.g., "Example PM
— Grad Adj Annual PM25 Spatial Field 2020.csv"). The table below describes the
variables in the output file.
Output file name: "Scenario Name + Grad Adj Annual PM25 Spatial Field"
Variable
Jd
gridcelljat
gridcelljong
b_pm25_ann_DV_ga
f_pm25_ann_DV_ga
b_blank_mass_ga
b_crusta l_mass_g a
b_EC_mass_ga
b_NH4_mass_ga
b_Ocmb_mass_ga
b_S04_mass_ga
b_N03_mass_ga
b_water_mass_ga
b_salt_mass_ga
f_blank_mass_ga
f_crustal_mass_ga
f_EC_mass_ga
f_NH4_mass_ga
f_Ocmb_mass_ga
f_S04_mass_ga
f_N03_mass_ga
f_water_mass_ga
f_salt_mass_ga
Description
The ID is a unique identifier for each model grid cell. The default value is the
column identifier multiplied by 1000 plus the row. (This is a character
variable.)
Latitude at the grid cell centroid in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are negative.
Longitude at the grid cell centroid in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g., United
States) are negative.
Base year gradient adjusted interpolated PM2.5 from unofficial data file
Future year gradient adjusted interpolated PM2.5
Base year blank mass concentration (ug/m3)
Base year crustal mass concentration (ug/m3)
Base year elemental carbon mass concentration (ug/m3)
Base year ammonium mass concentration (ug/m3)
Base year organic carbon mass (by difference) concentration (ug/m3)
Base year sulfate ion mass concentration (ug/m3)
Base year nitrate ion mass concentration (ug/m3)
Base year water mass concentration (ug/m3)
Base year salt mass concentration (ug/m3)
Future year blank mass concentration (ug/m3)
Future year crustal mass concentration (ug/m3)
Future year elemental carbon mass concentration (ug/m3)
Future year ammonium mass concentration (ug/m3)
Future year organic carbon mass (by difference) concentration (ug/m3)
Future year sulfate ion mass concentration (ug/m3)
Future year nitrate ion mass concentration (ug/m3)
Future year water mass concentration (ug/m3)
Future year salt mass concentration (ug/m3)
Abt Associates Inc.
106
June 2010
-------
Annual PM Analysis: Details
rrf crustal
Resultant annual relative response factor- Crustal Mass
Resultant annual response factor- Elemental Carbon Mass
Resultant annual relative response factor- Ammonium Mass.
Resultant annual relative response factor- Organic Carbon Mass
rrf ec
rrf nh4
rrf oc
rrf so4
Resultant annual relative response factor- Sulfate Mass
Resultant annual relative response factor- Nitrate Mass
Resultant annual relative response factor- Water Mass
Resultant annual relative response factor- Salt Mass
rrf no3
rrf water mass
rrf salt
5.2.2 Miscellaneous Output
MATS gives you ability to generate a number of specialized files for quality assurance,
identify the monitor in each county with the highest forecast, and other reasons. These
files include:
Quarterly Average Files. Contain baseline PM2.5 and species concentrations, RRFs,
forecasted PM2.5 and species concentrations. Available for point estimates and spatial
fields, as well as gradient-adjusted spatial fields. These files are important because all of
the basic PM2.5 calculations in MATS occur on a quarterly average basis. The "true"
RRFs and PM2.5 and species concentrations are found in the quarterly average files.
High County Sites. Monitor chosen in county based on highest future-year value.
Contains baseline PM2.5 and species concentrations, RRFs, forecasted PM2.5 and
species. File only available for the Standard Analysis (point) results.
Species Fractions for Spatial Fields. Contains quarterly species fractions and
interpolated species values (with or without gradient-adjustment) for spatial fields.
Quarterly Average Speciated Monitors. Contains baseline quarterly average monitored
species concentrations at STN and IMPROVE sites. This is the subset of species data
that MATS uses for each particular scenario (based on the MATS inputs and
configuration settings).
Neighbor Files. Contains neighbor identifiers for interpolation to both FRM sites or to
spatial fields.
The Quarterly Average Files provide intermediate calculations performed by MATS. In
particular, these files have the weighted quarterly average baseline and future values for
PM2.5 and its constituent species. In addition it gives the speciated RRFs. There are
potentially three Quarterly Average Files: one for Point Estimates and two for Spatial Field
Estimates (with and without gradient adjustment).
5.2.2.1 Quarterly Average Files
Abt Associates Inc.
June 2010
107
-------
Annual PM Analysis: Details
Output Description - Quarterly Average Point
The output file is named "Quarterly PM25 Point" with the Scenario Name appended at
the beginning (e.g., "Example Annual PM — Quarterly PM25 Point.csv"). The table
below describes the variables in the output file.
Output file name:
Variable
_id
Jype
_state_name
_county_name
monitorjat
monitorjong
monitor_gridcell
quarter
b_pm25_ann_q_DV
f_pm25_ann_q_DV
b_blank_mass_q
b_crustal_mass_q
b_EC_mass_q
b_NH4_mass_q
b_Ocmb_mass_q
b_S04_mass_q
b_N03_mass_q
b_water_mass_q
f_blank_mass_q
f_crustal_mass_q
f_EC_mass_q
f_NH4_mass_q
f_Ocmb_mass_q
f_S04_mass_q
f_N03_mass_q
f_water_mass_q
rrf_crustal
rrf_ec
rrf nh4
"Scenario Name + Quarterly PM25 Point + Year"
Description
The ID is a unique name for each monitor in a particular location.
The default value is the AIRS ID. (This is a character variable.)
FRM data
State name. (This is a character variable.)
County name. (This is a character variable.)
Latitude at the monitor site in decimal degrees. Values in the
northern hemisphere are positive, and those in the southern
hemisphere are negative.
Longitude at the monitor site in decimal degrees. Values in the
eastern hemisphere are positive, and those in the western
hemisphere (e.g., United States) are negative.
Identifier of grid cell closest to the monitor
Quarter
Base year 5 year weighted average PM2.5 (quarter) design
value
Future year 5 year weighted average PM2.5 (quarter) design
value
Base year blank mass concentration (ug/m3)
Base year crustal mass concentration (ug/m3)
Base year elemental carbon mass concentration (ug/m3)
Base year ammonium mass concentration (ug/m3)
Base year organic carbon mass (by difference) concentration
(ug/m3)
Base year sulfate ion mass concentration (ug/m3)
Base year nitrate ion mass concentration (ug/m3)
Base year water mass concentration (ug/m3)
Future year blank mass concentration (ug/m3)
Future year crustal mass concentration (ug/m3)
Future year elemental carbon mass concentration (ug/m3)
Future year ammonium mass concentration (ug/m3)
Future year organic carbon mass (by difference) concentration
(ug/m3)
Future year sulfate ion mass concentration (ug/m3)
Future year nitrate ion mass concentration (ug/m3)
Future year water mass concentration (ug/m3)
Relative response factor- Crustal Mass
Relative response factor- Elemental Carbon Mass
Relative response factor- Ammonium Mass. (Only calculated if
ammonium in model data and measured ammonium ["NH4"]
used in calculations.)
Abt Associates Inc.
108
June 2010
-------
Annual PM Analysis: Details
rrf_oc Relative response factor- Organic Carbon Mass
rrf_so4 Relative response factor- Sulfate Mass
rrf_no3 Relative response factor- Nitrate Mass
Output Description - Quarterly Average Spatial Field
The output file is named "Quarterly PM25 Spatial Field' with the Scenario Name
appended at the beginning (e.g., "Example Annual PM — Quarterly PM25 Spatial Field,
csv"). The table below describes the variables in the output file.
Output file name:
Variable
_id
gridcelljat
gridcelljong
quarter
b_pm25_ann_q_DV
f_pm25_ann_q_DV
b_blank_mass_q
b_crustal_mass_q
b_EC_mass_q
b_NH4_mass_q
b_Ocmb_mass_q
b_S04_mass_q
b_N03_mass_q
b_water_mass_q
f_blank_mass_q
f_crustal_mass_q
f_EC_mass_q
f_NH4_mass_q
f_Ocmb_mass_q
f_S04_mass_q
f_N03_mass_q
f_water_mass_q
rrf_crustal
rrf_ec
rrf nh4
'Scenario Name + Quarterly PM25 Spatial Field + Year"
Description
The ID is a unique identifier for each model grid cell. The default
value is the column identifier multiplied by 1000 plus the row.
(This is a character variable.)
Latitude at the grid cell centroid in decimal degrees. Values in the
northern hemisphere are positive, and those in the southern
hemisphere are negative.
Longitude at the grid cell centroid in decimal degrees. Values in
the eastern hemisphere are positive, and those in the western
hemisphere (e.g., United States) are negative.
Quarter
Base year 5 year weighted average PM2.5 (quarter) design value
Future year 5 year weighted average PM2.5 (quarter) design
value
Base year blank mass concentration (ug/m3)
Base year crustal mass concentration (ug/m3)
Base year elemental carbon mass concentration (ug/m3)
Base year ammonium mass concentration (ug/m3)
Base year organic carbon mass (by difference) concentration
(ug/m3)
Base year sulfate ion mass concentration (ug/m3)
Base year nitrate ion mass concentration (ug/m3)
Base year water mass concentration (ug/m3)
Future year blank mass concentration (ug/m3)
Future year crustal mass concentration (ug/m3)
Future year elemental carbon mass concentration (ug/m3)
Future year ammonium mass concentration (ug/m3)
Future year organic carbon mass (by difference) concentration
(ug/m3)
Future year sulfate ion mass concentration (ug/m3)
Future year nitrate ion mass concentration (ug/m3)
Future year water mass concentration (ug/m3)
Relative response factor- Crustal Mass
Relative response factor- Elemental Carbon Mass
Relative response factor- Ammonium Mass. (Only calculated if
ammonium in model data and measured ammonium ["NH4"]
used in calculations.)
Abt Associates Inc.
109
June 2010
-------
Annual PM Analysis: Details
rrf_oc Relative response factor- Organic Carbon Mass
rrf_so4 Relative response factor- Sulfate Mass
rrf_no3 Relative response factor- Nitrate Mass
Output Description - Quarterly Average Spatial Field - Gradient Adjusted
The output file is named "GrcidAdj Quarterly PM25 Spatial Field' with the Scenario
Name appended at the beginning (e.g., "Example Annual PM — Grad Adj Quarterly PM25
SpatialField.csv"). The table below describes the variables in the output file.
Output file name: "Scenario Name + Grad Adj Quarterly PM25 Spatial Field + Year"
Variable
id
gridcelljat
gridcelljong
quarter
b_pm25_ann_q_DV_ga
f_pm25_ann_q_DV_ga
b_blank_mass_q_ga
b_crusta l_m ass_q_g a
b_EC_mass_q_ga
b_NH4_mass_q_ga
b_Ocm b_m ass_q_g a
b_S04_mass_q_ga
b_N03_mass_q_ga
b_water_mass_q_ga
f_blank_mass_q_ga
f_crusta l_m ass_q_g a
f_EC_mass_q_ga
f_NH4_mass_q_ga
f_Ocmb_mass_q_ga
f_S04_mass_q_ga
f_N03_mass_q_ga
f_wate r_m ass_q_g a
rrf_crustal_ga
rrf_ec_ga
rrf_nh4_ga
Description
The ID is a unique identifier for each model grid cell. The default
value is the column identifier multiplied by 1000 plus the row.
(This is a character variable.)
Latitude at the grid cell centroid in decimal degrees. Values in the
northern hemisphere are positive, and those in the southern
hemisphere are negative.
Longitude at the grid cell centroid in decimal degrees. Values in
the eastern hemisphere are positive, and those in the western
hemisphere (e.g., United States) are negative.
Quarter
Base year 5 year weighted average PM2.5 (quarter) design value
Future year 5 year weighted average PM2.5 (quarter) design
value
Base year blank mass concentration (ug/m3)
Base year crustal mass concentration (ug/m3)
Base year elemental carbon mass concentration (ug/m3)
Base year ammonium mass concentration (ug/m3)
Base year organic carbon mass (by difference) concentration
(ug/m3)
Base year sulfate ion mass concentration (ug/m3)
Base year nitrate ion mass concentration (ug/m3)
Base year water mass concentration (ug/m3)
Future year blank mass concentration (ug/m3)
Future year crustal mass concentration (ug/m3)
Future year elemental carbon mass concentration (ug/m3)
Future year ammonium mass concentration (ug/m3)
Future year organic carbon mass (by difference) concentration
(ug/m3)
Future year sulfate ion mass concentration (ug/m3)
Future year nitrate ion mass concentration (ug/m3)
Future year water mass concentration (ug/m3)
Relative response factor- Crustal Mass
Relative response factor- Elemental Carbon Mass
Relative response factor- Ammonium Mass. (Only calculated if
ammonium in model data and measured ammonium ["NH4"] used
in calculations.)
Abt Associates Inc.
110
June 2010
-------
Annual PM Analysis: Details
rrf_oc_ga Relative response factor- Organic Carbon Mass
rrf_so4_ga Relative response factor- Sulfate Mass
rrf_no3_ga Relative response factor- Nitrate Mass
5.2.2.2 Output Description - High County Sites
In this file, MATS reports the monitor with the highest forecasted PM2.5 design value in
each county. The name of this file is "High County Sites" with the Scenario Name
appended at the beginning (e.g., "Example Annual PM — High County Sites.csv"). The
table below describes the variables in the output file.
Output file name:
Variable
Jd
Jype
_state_name
_county_name
monitor lat
monitorjong
monitor_gridcell
b_pm25_ann_DV
f_pm25_ann_DV
b_blank_mass
b_crustal_mass
b_EC_mass
b_NH4_mass
b_Ocmb_mass
b_S04_mass
b_N03_mass
b_water_mass
f_blank_mass
f_crustal_mass
f_EC_mass
f_NH4_mass
f_Ocmb_mass
f_S04_mass
f N03 mass
'Scenario Name + High County Sites + Year"
Description
The ID is a unique name for each monitor in a particular location. The
default value is the AIRS ID. (This is a character variable.)
FRM data
State name. (This is a character variable.)
County name. (This is a character variable.)
Latitude at the monitor site in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are
negative.
Longitude at the monitor site in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g.,
United States) are negative.
Identifier of grid cell closest to the monitor
Base year 5 year weighted average PM2.5 annual design value
Future year 5 year weighted average PM2.5 annual design value
Base year blank mass concentration (ug/m3)
Base year crustal mass concentration (ug/m3)
Base year elemental carbon mass concentration (ug/m3)
Base year ammonium mass concentration (ug/m3)
Base year organic carbon mass (by difference) concentration (ug/m3)
Base year sulfate ion mass concentration (ug/m3)
Base year nitrate ion mass concentration (ug/m3)
Base year water mass concentration (ug/m3)
Future year blank mass concentration (ug/m3)
Future year crustal mass concentration (ug/m3)
Future year elemental carbon mass concentration (ug/m3)
Future year ammonium mass concentration (ug/m3)
Future year organic carbon mass (by difference) concentration (ug/m3)
Future year sulfate ion mass concentration (ug/m3)
Future year nitrate ion mass concentration (ug/m3)
Abt Associates Inc.
111
June 2010
-------
Annual PM Analysis: Details
f water mass
Future year water mass concentration (ug/m3)
5.2.2.3 Species Fractions Spatial Field
Contains quarterly species fractions and interpolated species values (with or without
gradient-adjustment) for spatial fields. (Note that you cannot currently load this file into
MATS, and instead this file is currently only for information. A future version of MATS
may allow loading in species fractions for spatial fields.)
Output Description - Species Fractions Spatial Field
In this file, MATS reports the quarterly species fractions. The name of this file is "
Quarterly Avg Spec Frac Spatial Field' with the Scenario Name appended at the
beginning (e.g., "Example Annual PM — Quarterly Avg Spec Frac Spatial Field.csv"). The
table below describes the variables in the output file.
Output file name: "Scenario Name + Quarterly Avg Spec Frac Spatial Field"
Variable Description
ID
The ID is a unique identifier for each model grid cell. The
default value is the column identifier multiplied by 1000
plus the row. (This is a character variable.)
Latitude at the grid cell centroid in decimal degrees. Values
in the northern hemisphere are positive, and those in the
southern hemisphere are negative.
Longitude at the grid cell centroid in decimal degrees.
Values in the eastern hemisphere are positive, and those
in the western hemisphere (e.g., United States) are
negative.
Quarter
PM2.5 mass used to calculate species fractions (calculated
from the "PM2.5 for fractions" file
Crustal fraction of PM2.5 mass
gridcelljat
gridcelljong
quarter
PM25 mass frac
fcr
fee
fnh4
focm
fso4
fno3
Ammonium fraction of PM2.5 mass
Nitrate ion fraction of PM2.5 mass
Water fraction of PM2.5 mass
Salt fraction of PM2.5 mass
Blank mass
Organic carbon fraction of PM2.5 mass
Sulfate ion fraction of PM2.5 mass
Elemental carbon fraction of PM2.5 mass
fwater
fsalt
blank_mass
don
Degree of neutralization of sulfate used to calculate
ammonium mass (0.000 - 0.375)
Interpolated sulfate ion
S04
N03R
OCB
EC
Interpolated nitrate ion
Interpolated blank-adjusted organic carbon
Interpolated elemental carbon
CRUSTAL
DON
Interpolated crustal
Interpolated degree of neutralization of sulfate (DON).
Abt Associates Inc.
112
June 2010
-------
Annual PM Analysis: Details
i_SALT
i NH4
Interpolated salt
Interpolated ammonium
Output Description - Gradient-Adjusted Species Fractions Spatial Field
In this file, MATS reports the gradient adjusted quarterly species fractions. The name of
this file is "Grcid Adj Quarterly Avg Spec Frac Spatial Field' with the Scenario Name
appended at the beginning (e.g., "Example Annual PM — Grad Adj Quarterly Avg Spec
Frac SpatialField.csv"). The table below describes the variables in the output file.
Output file name: '
Field"
Variable
JD
gridcelljat
gridcelljong
quarter
PM25_mass_frac_ga
fcr_ga
fec_ga
fnh4_ga
focm_ga
fso4_ga
fno3_ga
fwater_ga
fsalt_ga
blank_mass_ga
don_ga
_S04_ga
_N03R_ga
_OCB_ga
_EC_ga
_CRUSTAL_ga
.DON
i_SALT_ga
i_NH4_ga
'Scenario Name + Grad Adj Quarterly Avg Spec Frac Spatial
Description
The ID is a unique identifier for each model grid cell. The
default value is the column identifier multiplied by 1000 plus
the row. (This is a character variable.)
Latitude at the grid cell centroid in decimal degrees. Values
in the northern hemisphere are positive, and those in the
southern hemisphere are negative.
Longitude at the grid cell centroid in decimal degrees.
Values in the eastern hemisphere are positive, and those in
the western hemisphere (e.g., United States) are negative.
Quarter
PM2.5 mass used to calculate species fractions (calculated
from the "PM2.5 for fractions" file)
Crustal fraction of PM2.5 mass
Elemental carbon fraction of PM2.5 mass
Ammonium fraction of PM2.5 mass
Organic carbon fraction of PM2.5 mass
Sulfate ion fraction of PM2.5 mass
Nitrate ion fraction of PM2.5 mass
Water fraction of PM2.5 mass
Salt fraction of PM2.5 mass
Blank mass
Degree of neutralization of sulfate used to calculate
ammonium mass (0.000 - 0.375)
Interpolated sulfate ion
Interpolated nitrate ion
Interpolated blank-adjusted organic carbon
Interpolated elemental carbon
Interpolated crustal
Interpolated degree of neutralization of sulfate (DON).
Note that DON is not gradient-adjusted.
Interpolated salt
Interpolated ammonium
Abt Associates Inc.
113
June 2010
-------
Annual PM Analysis: Details
5.2.2.4 Output Description - Quarterly Average Speciated Monitors
In this file, MATS reports the quarterly averages at the speciated monitors. The name of
this file is "Quarterly Avg Speciated Monitors" with the Scenario Name appended at the
beginning (e.g., "Example Annual PM — Quarterly Avg Speciated Monitors.csv"). The
table below describes the variables in the output file.
Output file name:
Variable
Jd
Jype
_state_name
_county_name
monitorjat
monitorjong
monitor_gridcell
quarter
b_crustal_mass
b_EC_mass
b_NH4_mass
b_Ocmb_mass
b_S04_mass
b_N03_mass
b water mass
'Scenario Name + Quarterly Avg Speciated Monitors"
Description
IMPROVE/STN Site Code
State name. (This is a character variable.)
County name. (This is a character variable.)
Latitude at the monitor site in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are
negative.
Longitude at the monitor site in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g., United
States) are negative.
Identifier of grid cell closest to the monitor
Quarter
Base year crustal mass concentration (ug/m3)
Base year elemental carbon mass concentration (ug/m3)
Base year ammonium mass concentration (ug/m3)
Base year organic carbon mass (by difference) concentration (ug/m3)
Base year sulfate ion mass concentration (ug/m3)
Base year nitrate ion mass concentration (ug/m3)
Base year water mass concentration (ug/m3)
5.2.2.5 Neighbor Files
MATS calculates the nearby monitors or "neighbors" separately for each species when
interpolating to points, and similarly for interpolating to spatial fields, which also involves
interpolation of PM monitors.
Neighbors for Interpolating Species to Points.
Neighbors for Interpolating PM2.5 & Species to Spatial Fields.
Output Description - Neighbors for Interpolating Species to Points
In this file, MATS reports the neighbors involved in interpolating species values (except
NH4 and DON) to the FRM monitor sites. The name of this file is "Neighbor File Point"
with the Scenario Name appended at the beginning (e.g., "Example Annual PM—
Neighbor File Point.csv"). The table below describes the variables in the output file.
Output file name: "Scenario Name + Neighbor File Point"
Abt Associates Inc.
114
June 2010
-------
Annual PM Analysis: Details
Variable
Jd
_state_name
_county_name
monitorjat
monitorjong
monitor_gridcell
quarter
_neighbor
neighbor_gridcell
distance
weightdistance
weightdistancesquared
pollutant
Description
The ID is a unique name for each monitor in a particular location.
The default value is the AIRS ID. (This is a character variable.)
State name. (This is a character variable.)
County name. (This is a character variable.)
Latitude at the monitor site in decimal degrees. Values in the
northern hemisphere are positive, and those in the southern
hemisphere are negative.
Longitude at the monitor site in decimal degrees. Values in the
eastern hemisphere are positive, and those in the western
hemisphere (e.g., United States) are negative.
Identifier of grid cell closest to the monitor
Quarter
IMPROVE/STN Site Code. (This is a character variable.)
Identifier of grid cell closest to the neighbor
Distance in kilometers from FRM monitor site and
IMPROVE/STN/SPECTRE neighbor.
Inverse-distance weight
Inverse-distance-squared weight
Pollutant (e.g., S04). Note interpolation approach can vary by
pollutant
Output Description - Neighbors for Interpoatling PM & Species to Spatial Field
In this file, MATS reports the neighbors involved in interpolating NH4 and DON to the
Spatial Field. The name of this file is "Neighbor File Spatial Field - PM' with the
Scenario Name appended at the beginning (e.g., "Example Annual PM — Neighbor File
Spatial Field - PM.csv"). The table below describes the variables in the output file.
Output file name:
Variable
Jd
gridcelljat
gridcelljong
quarter
_neighbor
neighbor_gridcell
distance
weight
Interpolation method
'Scenario Name + Neighbor File Spatial Field"
Description
The ID is a unique identifier for each model grid cell. The default
value is the column identifier multiplied by 1000 plus the row. (This
is a character variable.)
Latitude at the grid cell centroid in decimal degrees. Values in the
northern hemisphere are positive, and those in the southern
hemisphere are negative.
Longitude at the grid cell centroid in decimal degrees. Values in
the eastern hemisphere are positive, and those in the western
hemisphere (e.g., United States) are negative.
Quarter
IMPROVE/STN Site Code. (This is a character variable.)
Identifier of grid cell closest to the neighbor
Distance in kilometers from FRM monitor site and
IMPROVE/STN/SPECTRE neighbor.
Inverse-distance weight
Inverse-distance-squared weight
Abt Associates Inc.
115
June 2010
-------
Annual PM Analysis: Details
pollutant Pollutant (e.g., S04). Note interpolation approach can vary by
pollutant
5.3 Data Input
In the Data Input window, you specify the MATS input files that are used in each scenario.
There are three main types of files which must be specified. These include ambient
PM2.5 species data, ambient total PM2.5 data (FRM and IMPROVE), and gridded model
output data (e.g. CMAQ or CAMx data).
There is specific terminology that is used on the Data Input page. "Official" data refers to
PM2.5 FRM data that can be used to determine official design values for compliance
purposes (comparison to the NAAQS). Other datasets which may not have rigid regulatory
significance are sometimes referred to as "unofficial" data. The individual input file
choices are explained below.
Annual PM Analysis
]
¦ Choose Desired Output
¦ Output Choices-Advanced
Data Input
¦ Species Fractions Options
¦ Species Fractions - Advanced
¦ PM 2.5 Calculation 0 ptions
¦ Model Data Options
¦ Final Output and Check
Data Input
Species Data
® IS cedes MonitorD ata File! | C: \Program FilesSAbt Associates\M AT S \S ampleD ata\S pecies-f
O Species Fractions File | point
| spatial field ^
PM2.5 Monitor Data
Unofficial Dailv Averaae PM2.5 Data File ffor All Species Fractions & PM2.5 Spatial Field!
| C: \Program FilesSAbt AssociatesSM AT S \S ampleD ataSPM 25-f or-f ractions-020G-v2. csv
Official Quarterly Average FRM Data File (for PM2.5 Point Calculations)
| C: \Program FilesSAbt AssociatesSM AT S SS ampleD ataSAnnual-official-FR M -99-07-v2. csv
Model Data
® Daily model data input O Quarterly model data input
B aseline File | C: SProgram FilesSAbt AssociatesSM AT S SS ampleD ataS2002cc_E U S_PM 25_sub. csv
Forecast File |C:\Program FilesSAbt AssociatesSMATSSSampleDataS2020cc_EUS_PM25_sub.csv
< Back Next > Cancel
Species Data. MATS needs ambient PM2.5 species data to calculate species
concentrations at FRM monitoring sites and spatial fields. Users have a choice of
supplying a "Species Monitor Data File" or a "Species Fractions File".
Species Monitor Data File. The default is to provide a species monitor data file.
MATS is populated with daily average species data from STN and IMPROVE sites
across the country. However, users can also provide their own ambient data file.
MATS uses the daily average species data to calculate species fractions at each FRM
monitor (or at each grid cell, in the case of spatial fields). The species fraction data is
Abt Associates Inc.
116
June 2010
-------
Annual PM Analysis: Details
combined with the "unofficial daily average PM2.5 data" to calculate species
concentrations. The default MATS species data file contains all available data.
However, there is a data flag to indicate site days that are recommended to be removed
from the species fractions calculations. This is not necessarily the same data flags that
have been identified by State agencies. We have incorporated flagging routines that
remove data that are considered to be outliers and/or incomplete data. (A description
of the flags is provided in the section on Species Fractions Calculation Options.)
Species Fraction File. Alternatively, the user can choose to use a pre-calculated
species fractions file which contains quarterly species information for the FRM
monitors of interest. MATS can also re-use a species fractions file (either "point" or
"spatial fields") that has previously been generated by MATS. To re-use a previously
created fractions file, simply supply the correct path to the file. When re-using spatial
fields species fractions files, either "spatial field" or gradient adjusted spatial fields"
must be selected from the drop-down box. MATS cannot use both types of spatial
fields species fractions files at the same time . (The calculation of species fractions is
discussed here.)
PM2.5 Monitor Data. MATS uses both "official" and "unofficial" data in its
calculations.
Unofficial Daily Average PM2.5 Data File. The "unofficial daily average PM2.5" file
contains the PM2.5 data that is needed to calculate species fractions. It is used in
combination with the "species monitor data file" from above. The unofficial daily
average PM2.5 file is not needed if the user supplies a pre-calculated species fractions
file.
Similar to the species monitor data file from above, the "unofficial" PM2.5 data file
contains a data flag to indicate site days that are recommended to be removed from the
species fractions calculations. The flagged data is matched between the species file and
the PM2.5 file so that the same site days are removed. However, the PM2.5 data file
contains additional data (sites that don't contain speciation measurements) and therefore
has additional flagged site days. These are not the same data flags that have been
identified by State agencies. We have incorporated flagging routines that remove data
that are considered to be outliers and/or incomplete data. (A description of the flags is
provided in the section on Species Fractions Calculation Options.) The user is free to
unflag existing data or add flags as necessary and appropriate.
Official Quarterly Average FRM Data File. The "official quarterly average file"
contains all of the "official" quarterly average FRM data that has been used to
calculate PM2.5 design values. It is used to calculate design values and 5 year
weighted average design values as part of the attainment test.
The default data file in MATS was created by the Air Quality Analysis Group within
OAQPS. In most cases, the data should not be altered, however in some cases (e.g.
Abt Associates Inc.
117
June 2010
-------
Annual PM Analysis: Details
sensitivity analyses) there may be a need to add or remove data.
Model Data The "model data" refers to gridded model output from models such as
CMAQ or CAMx. The user can choose either daily model data input or quarterly model
data input (which is just a quarterly average of the daily model data). Either will work
for annual PM2.5. The default setting is daily average data. Recall that MATS can
generate quarterly average model data (which can then be re-used in subsequent MATS
runs).
Model data must be selected for all MATS runs. The size of the modeling grid defines
the outputs for point estimates and for spatial fields. For point estimates, MATS will
output the results for all specified monitors within the domain. For spatial fields, MATS
will create spatial fields that match the size of the gridded model domain.
Note that you need to specify both a Baseline File and a Forecast File. The baseline file
should be consistent with the historical monitor data that you use, and the forecast year
is the future-year of interest.
5.3.1 Species Data Input
The species data may be in form of monitor data (specified below) that MATS then uses to
calculate species fractions, or it may be in the form of species fractions directly (specified
here).
Monitor data should be in the form of a simple text file. The first row specifies the
frequency of the data (e.g., day). The second row presents comma-separated variable
names. The third row begins the data values. Below is an example of the monitor data file
format and descriptions of the variables in the file.
Format of Speciated Monitor Data
Abt Associates Inc.
118
June 2010
-------
Annual PM Analysis: Details
Species-for-fractions-0206-v2.csv - WordPad
File Edit View Insert Format Help
De?H # & £4 & ^ ffl ^ %
Courier New
v 10 v Western
Day
_id,lat,long,
"010050002", 3
"010050002",3
"010050002",3
"010050002",3
"010050002",3
"010050002",3
"010050002",3
"010050002",3
"010050002",3
"010050002",3
"010050002",3
"010050002",3
"010050002",3
"010050002",3
"010050002",3
_TYP E,D ATE, S 04, W03
1.664137,-85.60623
1.664137,-65.60623
1.664137,-35.60623
1.664137,-85.60623
1.664137,-85.60623
1.664137,-85.60623
1.664137,-85.60623
1.664137,-85.60623
1.664137,-85.60623
1.664137,-85.60623
1.664137,-85.60623
1.664137,-85.60623
1.664137,-85.60623
1.664137,-85.60623
1.664137,-85.60623
R, NH4, OCB, EC, CRUSTAL, 3ALT, DON, H20_AIM, OC, W03 , S04_3 3, C]
4,"STW",2 003100 6,7.21,0,2.8,6.92,0.42,0.55,0,0.375,2.!
4,"STN",20031012,5.99,0,2.16,,,0.06,0.05,0.361,2.22,,!
4,"STW",2 0031018,4.68,0,1.75,5.58,0.55,0.19,0,0.374,1
4,"STW",2 0031024,3.47,0,1.1,7.12,0.38,0.78,0,0.317,1.:
4,"STW",2 0031030,2.42, 0,0.74,3.7, 0.21,0.31,0.02,0.306.
4,"STW",20031105,2.39,0,0.48, 1.08,0.15,0.17, 0.02,0.20:
4,"STW",2 0031111,1.43,0,0.34,2.91,0.2,0.28,0,0.238,0.;
4,"STW",2 0031117,2.76,0,0.69, 1.8,0.2,0.3, 0.01, 0.25,0.;
4,"STW",20031123,3.41,0,1.21,3.23,0.5,0.24,0,0.355,1.;
4,"STW",2 003112 9, 1.3,0,0.47,0.9,0.12,0.18,0, 0.362,0.4!
4,"STW",2 00312 05,3.15,0, 1.13,2.17,0.13,0.19,0,0.3 59, 1
4,"STW",2 0031211,2.87,0, 1.03,0.69,0.13,0.09, 0,0.359, 1
4,"STW",2 0031217,1.88,0,0.76,0.3,0.13,0.06,0,0.375,O.i
4,"STW",20031223,2.27,0,0.7,1.59,0.37,0.09,0.02,0.308.
4,"STW",2 0031229,2.33,0,0.75,1.24,0.27,0.22,0.06,0.32;
irnmncnnn-:" ^ 1 fcdm rrciTMrr ^nrnm-ind
QdnriCTnm n^nnnnmn^-i r
For Help, press F1
NUM
Speciated Monitor Data Variable Descriptions
Variable
Jd
LAT
LONG
_TYPE
DATE
S04
N03R
NH4
OCB
EC
CRUSTAL
SALT
DON
H20 AIM
Description
The ID is a unique name for each monitor in a particular location. (This is a
character variable.)
Latitude at the monitor site in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are negative.
Longitude at the monitor site in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g., United
States) are negative.
STN or IMPROVE network data
Date of daily average ambient data with YYYYMMDD format (This is a
numeric variable)
Measured sulfate ion
Estimated nitrate retained on FRM filter, using measured nitrate, hourly T and
RH
Measured NH4+ ion
OC blank adjusted, STN uses constant blank value per sampler type,
IMPROVE uses backup filter methodology
Measured EC
Using IMPROVE algorithm, Crustal, aka "Fine Soil" = 2.2 * [Al] + 2.49 * [Si] +
1.63 x [Ca] + 2.42 x [Fe] + 1.94 x [Ti]
Estimated salt using CI (Salt=1,8*CI), where CI is elemental chloride
Degree of neutralization of S04 (0-0.375)
Calculated water using AIM and measured S04, adjusted N03 and measured
NH4
Abt Associates Inc.
119
June 2010
-------
Annual PM Analysis: Details
OC
N03
S04_3S
CRUSTAL_ALT
FRM_MASS
MEASURED_FM
RCFM
Al
Ca
Fe
Ti
Si
EPA_FLAG
USER FLAG
Measured OC
Measured nitrate ion
Sulfate value derived from S, as per IMPROVE protocol, i.e. S04_3S = 3*S.
Alternative Crustal calculation using measured Si, Fe, Ca, Ti (modified formula
without Al) = 3.73 x [Si] + 1.63 x [Ca] + 2.42 x [Fe] + 1.94 x [Ti]
FRM mass
STN or IMPROVE sampler measured fine mass (Teflon filter)
Reconstructed Fine Mass using IMPROVE protocol, RCFM = [Amm_Sulfate] +
[Amm_Nitrate] + [OCM] + [EC] + [Fine Soil]
Measured Al
Measured Ca
Measured Fe
Measured Ti
Measured Si
Flag to indicate data that EPA recommends to be removed from the species
fractions calculations. 0 = valid data, 1 or greater = data that has been flagged
and should be removed
Flag to indicate additional data that the user wants to remove from the species
fractions calculations. 0 = valid data, 1 or greater = data that has been flagged
and should be removed
Note:
Some variables are supplied for QA purposes only and are either not used by MATS or are
calculated internally by MATS. For example, "OC" is not used by MATS (OCb is used)
and H20AIM is calculated internally. Character variables have names that begin with an
underscore (i.e., "_"), and the character values used can be kept with or without quotes. (If
a character variable has an embedded space, such as might occur with the name of a
location, then use quotes.)
5.3.2 PM2.5 Monitor Data Input
MATS uses the "unofficial" daily PM2.5 file for the calculation of all species fractions
files (both point and spatial field) and in the calculation of PM2.5 levels in spatial fields.
MATS uses the "official" quarterly PM2.5 file for point estimates.
5.3.2.1 Unofficial Daily PM2.5 Monitor Data Input
Monitor data should be in the form of a simple text file. The first row specifies the
frequency of the data (e.g., day). The second row presents comma-separated variable
names. The third row begins the data values. Below is an example of the monitor data file
format and descriptions of the variables in the file.
Format of Unofficial PM2.5 Monitor Data
Abt Associates Inc.
120
June 2010
-------
Annual PM Analysis: Details
PM25-for-fractions-0206-v2.csv - WordPad
File Edit View Insert Format Help
~ ^ y a a #4 & mm %
Courier New
10
Western
B
U
|Day
_ID,_TYPE,LAT,LONG,
"010030010","FRH",3
"010030010","FRH",3
"010030010","FRH",3
"010030010","FRH",3
"010030010","FRH",3
"010030010","FRH",3
"010030010","FRH",3
"010030010","FRH",3
"010030010","FRH",3
"010030010","FRH",3
"010030010","FRH",3
"010030010","FRH",3
"010030010","FRH",3
"010030010","FRH",3
"010030010","FRH",3
"010030010","FRH",3
"n-inmnn-in" "fpw t
DATE,PH2 5,EPA FLAG,USER
0.498001046
0.498001046
0.498001046
0.498001046
0.498001046
0.498001046
0.498001046
0.498001046
0.498001046
0.498001046
0.498001046
0.498001046
0.498001046
0.498001046
0.498001046
0.498001046
n 4QRnnin4(=:
-87.88141232
-87.88141232
-87.88141232
-87.88141232
-87.88141232
-87.88141232
-87.88141232
-87.88141232
-87.88141232
-87.88141232
-87.88141232
-87.88141232
-87.88141232
-87.88141232
-87.88141232
-87.88141232
-R^ RR1419T?
FLAG
,20020102,
,20020105,
,20020108,
,20020111,
,20020114,
,20020117,
,20020120,
,20020123,
,20020126,
,20020129,
,20020201,
,20020204,
,20020207,
,20020210,
,20020213,
,20020216,
?nn?n?1q
12 .7,0,
12.5,0,
7.3,0,
6.7,0,
5.7,0,
9.9,0,
9,0,
9.7,0,
10.2,0,
9.8,0,
3.8,0,
5.4,0,
6.2,0,
10.2,0,
29,0,
13.8,0,
r 1 n
For Helpj press F1
NUM
Unofficial PM2.5 Monitor Data Variable Descriptions
Variable
JD
_TYPE
LAT
LONG
DATE
PM25
EPA_FLAG
USER FLAG
Description
The ID is a unique name for each monitor in a particular location. (This is
a character variable.)
FRM, IMPROVE, or other "special" data - may include dummy sites.
(This is a character variable.)
Latitude at the monitor site in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are
negative.
Longitude at the monitor site in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g.,
United States) are negative.
Date of daily average ambient data with YYYYMMDD format (This is a
numeric variable)
Measured PM2.5 mass (ug/m3)
Flag to indicate data that EPA recommends to be removed from the
species fractions calculations. 0 = valid data, 1 = data that has been
flagged and should be removed
Flag to indicate additional data that the user wants to remove from the
species fractions calculations. 0 = valid data, 1 or greater = data that has
been flagged and should be removed.
Note:
Character variables have names that begin with an underscore (i.e., and the character
values used can be kept with or without quotes. (If a character variable has an embedded
space, such as might occur with the name of a location, then use quotes.)
Abt Associates Inc.
121
June 2010
-------
Annual PM Analysis: Details
5.3.2.2 Official Quarterly PM2.5 Monitor Data Input
Monitor data should be in the form of a simple text file. The first row specifies the
frequency of the data (e.g., quarter). The second row presents comma-separated variable
names. The third row begins the data values. Below is an example of the monitor data file
format and descriptions of the variables in the file.
Format of Official Quarterly PM2.5 Monitor Data
Annual-official-FRM-99-Q7-v2.csv - WordPad
File Edit View Insert Format Help
~ e? S S & 1-4 Ji % 6,
Courier New
10
|Quarter
_ID, _TYPE,
"010030010"
"010030010"
"010030010"
"010030010"
"010030010"
"010030010"
"010030010"
"010030010"
"010030010"
"010030010"
"010030010"
"010030010"
"010030010"
"010030010"
"010030010"
"n 1 nm nn in»
LAT,
,"FRM'
,"FRMr
,"FRMr
,"FRMr
, "FRIT
,"FRH'
,"FRH'
,"FRH'
,"FRHr
,"FRMr
,"FRMr
, "FRIT
,"FRH'
,"FRH'
,"FRHr
LONG, DATE,
,30.49800104
,30.49800104
,30.49800104
,30.49800104
,30.49800104
,30.49800104
,30.49800104
,30.49800104
,30.49800104
,30.49800104
,30.49800104
,30.49800104
,30.49800104
,30.49800104
,30.49800104
PM25,
6,-87,
6,-87,
6,-87,
6,-87,
6,-87,
6,-87,
6,-87,
6,-87,
6,-87,
6,-87,
6,-87,
6,-87,
6,-87,
6,-87,
6,-87,
in jiannn 1 nne —p'?
WDAYS, _SUBSTITUTION_CODE, COMPLETION_CODE, _STATE_NAME, _COUNTY_NAME
,88141232,19990101,-9,-9,"",-9,"Alabama","Baldwin"
,88141232,19990401,-9,-9,"",-9,"Alabama","Baldwin"
, 88141232,19990701,-9,-9,"",-9,"Alabama","Baldwin"
,88141232,19991001,-9,-9,"",-9,"Alabama","Baldwin"
,88141232,20000101,11.81875,16,"",-9,"Alabama","Baldwin"
,88141232,20000401,13.24,25,"",-9,"Alabama","Baldwin"
,88141232,20000701,18.796,25,"",-9,"Alabama","Baldwin"
,88141232,20001001,14.257894737,19,"",-9,"Alabama","Baldwin"
,88141232,20010101,10.710714286,28,"",4,"Alabama","Baldwin"
,88141232,20010401,11.133333333,27,"",4,"Alabama","Baldwin"
,88141232,20010701,11.07,30,"",4,"Alabama","Baldwin"
,88141232,20011001,9.4032258065,31,"",4,"Alabama","Baldwin"
,88141232,2 0020101,9.71333 33 3 33,3 0,
,88141232,20020401,9.3137931034,29,
,88141232,20020701,12.665217391,23,
?nn?1nni a nQnitti1
maxq",2,"Alabama","Baldwin"
maxq",2,"Alabama","Baldwin"
maxq",2,"Alabama","Baldwin"
=or Helpj press F1
Official Quarterly PM2.5 Monitor Data Variable Descriptions
Variable Description
_ID The ID is a unique name for each monitor in a particular location. The
default value is the AIRS ID. (This is a character variable.)
_TYPE Leave blank
LAT Latitude at the monitor site in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are
negative.
LONG Longitude at the monitor site in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g.,
United States) are negative.
DATE Date of quarterly average ambient data with YYYYMMDD format. The
date represents the first day of each quarter. (This is a numeric variable)
PM25 PM2.5 mass
NDAYS Number of complete days in each quarter
_SUBSTITUTION_CODEIndicates whether the design value period was determined to be
complete by using substitution procedures.
COMPLETION_CODE Official design value completion codes (1, 2, 3, 4, or 5). Codes are valid
for the end year of each 3 year design value period.
_STATE_NAME State name. (This is a character variable.)
_COUNTY_NAME County name. (This is a character variable.)
Note:
Abt Associates Inc.
122
June 2010
-------
Annual PM Analysis: Details
Character variables have names that begin with an underscore (i.e., and the character
values used can be kept with or without quotes. (If a character variable has an embedded
space, such as might occur with the name of a location, then use quotes.)
5.3.3 Model Data Input
Model data should be in the form of a simple text file. The first row specifies the
frequency of the data (e.g., day). The second row presents comma-separated variable
names. The third row begins the data values. Below is an example of the model data file
format and descriptions of the variables in the file. Note that there is both a base year and
a future year model file. The format for both is the same.
Note that you can load in either daily model data or quarterly model data.
Model Data
® Daily model data input
Quarterly peak model data input
B aseline File | C: \Program Files V\bt Associates\MAT S \S ampleD ata\2002cc_E U S_PM 25_sub. cs v
Forecast File C:\Program Files\Abt Associates\MATS\SampleData\2020cc_EUS_PM25_sub.csv
...
...
Format of Model Data
2002cc_EUS_PM25_sub.csv - WordPad
File Edit View Insert Format Help
~ y # H #4 ^ n
b s u $) m m m :E
|day
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
ID,_TYPE,
65,"",32.
65, "
65, "
65, "
65, "
65, "
65, "
65, "
65, "
65, "
65, "
65, "
65, "
65, "
65, "
65, "
65, "
',32
',32
',32
',32
',32
',32
',32
',32
',32
',32
',32
',32
',32
',32
',32
',32
65,"",32
65,"",32
65, "
65, "
65, "
65, "
65, "
65, "
65, "
65, "
',32
',32
',32
',32
',32
',32
',32
',32
LAT
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
LONG ,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
DATE ,
20020101,
20020102,
20020103,
20020104,
20020105,
20020106,
20020107,
20020108,
20020109,
20020110,
20020111,
20020112,
20020113,
20020114,
20020115,
20020116,
20020117,
20020118,
20020119,
20020120,
20020121,
20020122,
20020123,
20020124,
20020125,
20020126,
20020127,
Crustal,
2.8938
9910
1356
2023
3665
4842
7300
8170
8733
9303
6944
4354
6101
4628
1597
6949
552 6
8016
2738
8043
9837
0872
6651
6103
2452
0805
NH4
0.8440
0064
3274
4508
1407
1455
1185
1678
1544
7045
8669
4571
1805
2882
6886
6052
4553
72 67
4397
9561
8742
0715
2512
3232
6444
1094
0487
S04
1.1525
2453
5503
2391
6652
3141
0372
0973
1099
2925
8649
6251
9503
0504
8693
9658
4824
6180
1787
8673
2578
3561
8760
2327
9694
0850
0349
EC
,3428
,3024
,3894
,5332
,4050
,0818
,2976
,4232
, 6474
,3994
,4779
,3200
,5070
,5513
,4941
,3923
, 6301
,3462
,2426
,3597
,3345
,4460
, 1455
, 1413
,2579
,5502
4185
8583
6868
4481
8896
1343
5331
6045
5588
9135
7046
7874
8407
8685
2 697
8483
9499
4840
1327
0177
3031
7230
0104
0067
9901
2805
2156
OC ,
,3510,
,1155,
,2507,
,0241,
,2245,
,4239,
,2708,
,7906,
,5978,
,2099,
,2104,
,7154,
,7200,
,8608,
,4518,
, 1626,
,7062,
,8798,
,2229,
, 6218,
,5093,
,7432,
,5435,
,4728,
,9955,
,5656,
,6647,
PM25 ,
8.3861,
8.8275,
11.6806,
13 .4746,
11.4404,
1.7175,
9.3578,
10.4348,
11.7966,
8.2250,
9.5399,
6.8945,
12.7001,
13.0609,
9.7304,
8.3897,
15.0897,
7.4988,
4.8955,
9.1079,
7.7430,
11.3870,
2.6622,
2.9357,
6.3848,
12 .4952,
13.1366,
4677
3373
4237
9857
5842
1237
4677
8712
7997
1342
7523
6504
0805
0743
0514
7299
102 6
6300
3462
6233
7217
7273
6255
8523
2 644
8330
6917
-or Helpj press F1
Model Data Variable Descriptions
Abt Associates Inc. June 2010
123
-------
Annual PM Analysis: Details
Variable
ID
_TYPE
LAT
LONG
DATE
CM
CRUSTAL
NH4
S04
EC
N03
OC
PM25
SALT
Note:
The "PM25" mass variable is used to calculate PM2.5 model "gradients" for gradient
adjusted spatial fields. It is up to the user to provide a modeled PM2.5 mass
concentration using an appropriate definition of modeled PM2.5 mass.
Character variables have names that begin with an underscore (i.e., "_"), and the
character values used can be kept with or without quotes. (If a character variable has an
embedded space, such as might occur with the name of a location, then use quotes.)
5.4 Species Fractions Calculation Options
The Species Fractions Calculation Options has two main sections. One involving
speciated monitor data (e.g., STN and IMPROVE monitors) and the other total PM2.5
monitor data (FRM and IMPROVE). For each type of data you can specify the years of
interest, whether you want to delete certain data, and the minimum amount of data for a
monitor to be considered "valid" (and thus included in the calculations). In the next
sections, we describe each of these three options in more detail. Note that these options
apply more or less in the same way for both speciated monitor data and total PM2.5
monitor data. Also note that these options are no longer relevant if you have loaded a
species fractions file.
IMPROVE-STN Monitor Data. The speciation data from STN and IMPROVE
monitors are interpolated by MATS in order to provide species data for any point in a
modeling domain. The interpolated species data is used to calculate species fractions at
FRM monitors (point estimates) and/or species fractions at all grid cells (spatial fields).
Note that you do not need to have values for all species for a monitor to be considered
valid, as each species is considered individually. However, the "EPA Flag" variable in
the default "species for fractions" file has been set so that all monitor days that do not
Description
The ID is a unique name for each monitor in a particular location. The default
value is the column identifier multiplied by 1000 plus the row. (This is a
character variable.)
Leave blank
Latitude at the grid cell centroid in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are negative.
Longitude at the grid cell centroid in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g., United
States) are negative.
Date of daily average model value with YYYYMMDD format (This is a numeric
variable)
Coarse PM mass (ug/m3)
Crustal PM2.5 mass
Ammonium mass
Sulfate PM
Elemental carbon
Nitrate PM
Organic mass PM
PM2.5 mass
Salt
Abt Associates Inc.
124
June 2010
-------
Annual PM Analysis: Details
have complete species data are not used in the calculations (flag =1). If the user wants
to use the incomplete species data, the flag can changed to "0".
PM2.5 Monitor Data. The total PM2.5 data from FRM are used by MATS to calculate
species fractions for point estimates (in conjunction with the interpolated speciation data
from STN and IMPROVE monitors). The interpolated species data is used to calculate
species fractions at FRM monitors (point estimates) and/or species fractions at all grid
cells (spatial fields).
Annual PM Analysis [
EH51 Choose Desired Output
Species Fractions Calculation Options
¦ Output Choices - Advanced
¦ Data Input
Species Fractions Options
IMPROVE-STN Monitor Data
¦ Species Fractions - Advanced
Monitor Data Years
¦ PM2.5 Calculation Options
Start Year End Year
¦ Model Data Options
12002 _rj 12004 _t]
Final Output and Check
Delete Specified Data Values
W EPA-specified deletions from monitor data
User-specified deletions from monitor data
Minimum Data Requirements
Minimum number of valid days per quarter
11±l
Minimum number of valid years required for valid season
Minimum number of valid seasons for valid monitor
PM2.5 Monitor Data
Monitor Data Years
Start Year End Year
12002 12004 _»J
Delete Specified Data Values
W EPA-specified deletions from monitor data
User-specified deletions from monitor data
Minimum Data Requirements
Minimum number of valid days per quarter
H±l
Minimum number of valid years required for valid season
Minimum number of valid seasons for valid monitor (point calculations)
Minimum number of valid seasons for valid monitor (spatial fields calculations) |
< Back
Next >
5.4.1 Monitor Data Years
Using the Start Year and End Year drop-down menu options, you can choose more than
one year of speciated data for the calculation of species fractions. The default approach in
MATS is to use three years of data.
Abt Associates Inc.
125
June 2010
-------
Annual PM Analysis: Details
Monitor Data Years
Start Year
End Year
2002
- [2004 V]
2003
2004
2005
200G
i Data Values
sletions from monitor data
Metions from monitor data
MATS handles multiple years of data by calculating averages for each species by quarter
and year. MATS then averages the quarterly values across the years (e.g., average quarter
1 values of S04 across two years to get a single "quarter 1" estimate). After completing
this step, MATS will have four quarterly estimates for each species at each speciated
monitor. These quarterly values are then ready to be interpolated to FRM sites or to grid
cell centroids in spatial fields.
The default is to delete the observations specified by EPA. (That is, these observations are
excluded from a particular analysis, while they of course remain in the database.) As
described in the Data Input section, valid data are given a value of "0" and observations
that are deleted are given a value of "1" to "10", as follows:
0. Data is OK to use
1. Important species is missing (e.g. no3, so4, oc, etc)
2. Constructed mass < 30% total mass
3. Constructed mass > 2 times total mass
4. Fire event
5. total mass < so4
6. total mass < crustal
7. OC is outside QA criteria
8. Soil is outside QA criteria
9. Both OC and soil are outside QA criteria
10. Regional concurrence on exceptional event
There is also an option for the user to flag data, using the same convention of "0" for valid
data and "1" to "10" for data marked for deletion. If both the EPA-specified and User-
specified flags are checked, then MATS deletes any observations that are marked for
deletion by either the EPA or the user. This makes it easy for the user to flag additional
data for removal from the calculations (without deleting the actual record from the ambient
data file).
5.4.2 Delete Specifed Data Values
Abt Associates Inc.
126
June 2010
-------
Annual PM Analysis: Details
Delete Specified Data Values
P" EPA-specified deletions from monitor data
~~ User-specified deletions from monitor data
5.4.3 Minimum Data Requirements
There are three sets of minimum data requirements:
Minimum number of valid days per valid quarter. This is the minimum number of
site-days per valid quarter. The default is 11 days, which corresponds to > 75%
completeness for monitors on a 1 in 6 day schedule. This is a minimum number of
samples that is routinely used in calculations of quarterly average concentrations.
Minimum number of valid quarters required for valid season. This is the number of
years of data (within the start year and end year specified) for which we have valid
quarters for a given season. The default value is 1 year. If the value is set = 2, then there
will need to be 2 years of valid data from quarterl in order for quarter one to be
considered complete (and the same for the other 3 quarters).
Minimum number of valid seasons required for valid monitor. This is the number of
valid seasons that are needed in order for a particular monitor's data to be considered
valid. The default is 1 for IMPROVE-STN monitor data and the range is 1-4. For
example, if the value is = 1, then a monitor's data will be used in the species fractions
calculations if it has at least one valid season. If the value = 4, then the site must have all
4 seasons of valid data to be used. The default for PM2.5 depends on whether the data
are used in point calculations (default = 4) or spatial field calculations (default =1).
Minimum Data Requirements
Minimum number of valid days per quarter | 11-$-|
Minimum number of valid years required for valid season
Minimum number of valid seasons for valid monitor (point calculations) i—«S
Minimum number of valid seasons for valid monitor (spatial fields calculations) i—m
Example 1: Minimum Days = 11, Minimum Years = 1, Minimum Seasons = 1
Consider the default assumptions and the following data from three monitors:
Abt Associates Inc.
127
June 2010
-------
Annual PM Analysis: Details
Monitor 1
Monitor 2
Monitor 3
Year
Quarter
?? Ohs.
Av<| <11 <1 in3)
?? Ohs.
Av<| (iiq/m3)
? Ohs.
Av<| (ufl/m3)
2002
1
8
10.2
11
10.8
11
10.4
2
6
13.3
13
13.2
8
13.7
3
11
10.7
4
11.2
12
12.9
4
5
12.4
12
13.7
8
11.5
2003
1
11
12.7
10
13.4
9
12.1
2
9
12.0
11
13.0
12
10.4
3
12
14.5
13
14.2
10
10.7
4
6
11.5
12
13.4
9
12.5
2004
1
6
13.8
15
14.2
12
14.9
2
7
14.1
11
14.5
10
12.0
3
12
14.9
12
14.6
10
10.1
4
13
12.6
9
12.1
12
14.1
With the default assumptions, MATS would then use the highlighted observations and
calculate averages for each quarter at each monitor in the following way:
Monitor 1
Monitor 2
Monitor 3
On tiller
Av
-------
Annual PM Analysis: Details
Monitor 1
Monitor 2
Monitor 3
Ouarter
Avg fug ni3)
Av(j (ug in3)
Avg fug in3)
1
--
12.5
12.6
2
--
13.6
--
3
12.5
14.4
--
4
--
13.6
--
Note that the requirement of 2 years, reduces to a single quarter the seasonal averages
calculated for Monitors 1 and 3. If you had further required that a monitor needed, say, 4
seasons, then MATS would have only calculated averages for Monitor 2.
5.5 Species Fractions Calculation Options - Advanced
The Species Fractions Calculation Options - Advanced screen allows you to make
relatively advanced choices for your analysis. Generally speaking, the default options
settings are consistent with the EPA modeling guidance document. One set of options
allows you to specify the interpolation weighting that you want to use and whether the
interpolation involves a maximum distance or not. The second set of options involves
choices regarding ammonium, blank mass, and organic carbon.
Annual PM Analysis
Choose Desired Output
Output Choices - Advanced
Species Fractions Calculation Options - Advanced
¦ Species Fractions Options
Species Fractions - Advanced
¦ PM2.5 Calculation Options
¦ Model Data Options
I^HFinal Output and Check
Interpolation Options
PM2.5 fi 350 Crustal |Inverse Distance Squared V ~ [90000 «] C »
S04 | Inverse Distance Squared V_^ 90000 Si DON | Inverse Distance Squared V ~ | 190000 ** y-j »
N03 | Inverse Distance Squared V ~ ] 190000 «|~vj »| OC | Inverse Distance Squared V 190000 »|
EC | Inverse Distance Squared V 190000 «[~^~| »| NH4
Salt Inverse Distance Squared V ~"[ 190000 «
Miscellaneous Options
Ammonium
® Use DON values
O Use measured ammonium
NH4 percentage evaporating (0-100)
I S3
Default Blank Mass
Default Blank Mass
Organic Carbon
Organic carbon mass balance floor f
Organic carbon mass balance ceiling
0 5
< Back
Next >
Cancel
5.5.1 Interpolation Options for Species Fractions Calculation
The Interpolation Options panel allows you to choose how you will interpolate, or
Abt Associates Inc.
129
June 2010
-------
Annual PM Analysis: Details
combine, the values from different monitors. One approach is to use Inverse Distance
Weights. This means that the weight given to any particular monitor is inversely
proportional to its distance from the point of interest. A second approach is Inverse
Distance Squared Weights, which means that the weights are inversely proportional to the
square of the distance. And the third approach is Equal Weighting of Monitors. The
default approach for PM is Inverse Distance Squared Weights.
Interpolation Options
[inverse Distance Squared \ T
19 0 0 0 0 0 «-$-»¦
Crustal [ Inverse Distance Squared *
|900000 *
: H
[inverse Distance Squared|
[900000 « »
DON [inverse Distance Squared S
|900000 «
Equal Weighting of Monitors
Inverse Distance Weiqhts
)000 « »
OC [inverse Distance Squared ^ T
900000 «
(inverse Distance Squared Weiqhts:
1000 « »
NH4
(900000 «-r- »
[inverse Distance Squared[
[900000 « C »
When interpolating monitor values, MATS allows you to identify the monitors you want to
use based on their distance away from the point of interest (e.g., the center of a grid cell).
The first step in the interpolation process is to identify the monitors that are nearby, or
neighbors, for each point of interest. The next step is to determine the distance (in
kilometers) from the nearby monitors to the point of interest.
The default approach is to include all valid monitors (i.e., those that satisfy the three
criteria in the Species Fractions Calculation Options panel), regardless of distance. If you
want to limit the use of monitors based on distance, type in the distance you want to use
(e.g., 100) next to the pollutant of interest.
Interpolation Options
Inverse Distance Squared ^~[ |900000_«
Inverse Distance Squared \
PM2.5
S04
NOx
EC
Salt
Inverse Distance Squared ^~| [900000 «
Inverse Distance Squared •*"] [900000 «
Inverse Distance Squared T | [900000 «
Crustal
DON
OC
NhM
Inverse Distance
Squared'' T
Inverse Distance Squared \
(900000 «
(900000 «
|900000 «
(900000 «
-
You can also change the number using the arrows. The double arrow on the right
increases the number in units of 100:
200
Jk.
T
and the double arrow on the left decreases the number in units of 100. The upper arrow
increases the number in single digits:
205
Jl
w
Abt Associates Inc.
130
June 2010
-------
Annual PM Analysis: Details
and the lower arrow reduces the number in single digits.
Note that a distance of one hundred (100) kilometers means that any monitors further than
100 kilometers can no longer be used in the interpolation. If a point of interest has no
monitors within the specified distance, then no value is calculated.
5.5.2 Miscellaneous Options
The Miscellaneous Options panel lets you make choices regarding:
Ammonium. This allows you to specify whether MATS uses degree of neutralization
(DON) values to calculate ammonium (NH4) or whether it uses measured ammonium in
conjunction with an assumption about the percentage of NH4 that evaporates. The
default option is to use DON values. If you want to use measured ammonium, you need
to click the button and choose a NH4 percentage evaporating (e.g., 50). The default is
"0", which assumes that no ammonium evaporates from the FRM filters. The
calculations underlying the default and alternative ammonium calculations are discussed
in detail in the section on species fractions calculations.
Default Blank Mass. The Default Blank Mass option simply allows you to set default
blank mass to the desired level. The default is 0.5. You can type desired value, or use
the arrows to increase or decrease the value.
Organic Carbon. This allows you to set the "floor" and the "ceiling" for the mass
balance calculation for organic carbon. The calculations involved are discussed in detail
in the section on species fractions calculations.
Miscellaneous Options
Ammonium
© Use DON values
O Use measured ammonium
NH4 percentage evaporating (0-100)
Default Blank Mass
Default Blank Mass | 0.5~^-|
Organic Carbon
Organic carbon mass balance floor | 1 vj
Organic carbon mass balance ceiling 0.8~^-|
5.5.3 Internal Precision of the Calculations
All calculations in MATS are carried out with single precision. In addition, most output
files by default generate outputs only up to 3 digits after the decimal. Therefore, the base
Abt Associates Inc.
131
June 2010
-------
Annual PM Analysis: Details
year DV (b_pm25_ann_q_DV) may differ slightly from the sum of the component species
of the quarterly PM2.5 MATS output files. This issue may be remedied by increasing
precision of the species and species fractions calculations from the default 3 significant
digits to 7 (or more) significant digits by modifying the MATS.ini file as follows: set
species_calc_precision=7 and species_fraction_precision=7. Please note that the future
year species always add up to the future year DV. However, increases in species fractions
precision may result in very small changes in future DV due to the dependence of the
future concentrations on the base year concentrations.
5.6 PM2.5 Calculation Options
The PM2.5 Calculation Options window allows you to specify the particular years of
monitor data that you want to use from the input file you specified in the Data Input
section. You can also specify the following:
Official vs. Custom Values. You can specify whether to use "official" design values,
which are generally recommended, or use "custom" design values. If you choose to use
custom design values, then you specify the minimum number of days of observations
each valid quarter and the minimum number of valid quarters.
Valid FRM Monitors. You can also specify the minimum number of design values (the
default is 1) and whether you want to make sure that particular design values have to be
used in the calculations.
NH4 Future Calculation. You can also specify how you want to forecast NH4 values.
The default approach is to use baseline DON values, and the alternative is to use baseline
NH4 and a RRF value for NH4.
Abt Associates Inc.
132
June 2010
-------
Annual PM Analysis: Details
Annual PM Analysis
Choose Desired Output
¦ Output Choices - Advanced
¦ Data Input
¦ Species Fractions Options
¦ Species Fractions - Advanced
PM2.5 Calculation Options
¦ Model Data Options
Final Output and Check
PM2.5 Calculation Options
PM2.5 Monitor Data Years
Start Year 2000 ^ End Year 2004
"31
® Official Desiqn Values
O Custom Desiqn Values
Valid FRM Quarters
Minimum days for valid quarter
Valid FRM Design Values
Minimum valid quarters in design value period 12
"3
Valid FRM Monitors
Minimum Number of Design Values
Required Design Values
F 3
-
-NH4 future calculation
(• Calculate future year NH4 using base year (constant) DON values
r Calculate future year N H 4 using base year N H 4 and the N H 4 R R F
< Back
Next >
5.6.1 PM2.5 Monitor Data Years
Using the Start Year and End Year drop-down menu options, you can choose more than
one year of official PM monitor data for the calculation of future PM2.5 values. The
default approach in MATS is to use five years of data.
PM2.5 Monitor Data Years
Start Year 12000
-
End Year 12004
J
1999
* Officii: Wlll'lli
r> ~ 2001
Custo 2002
5
Valid FP 2003
Minimum d. ^005
12006
I11 ±
MATS handles multiple years of data as described in the Baseline PM2.5 Calculation
section.
5.6.2 Design Values
MATS gives you option to choose "official" design values, which are generally
recommended, or to choose "custom" design values. If you choose to use custom design
values, then you need to specify the minimum number of days of observations each valid
quarter and the minimum number of valid quarters.
Minimum days for valid quarter. This is the minimum number of site-days per valid
Abt Associates Inc.
133
June 2010
-------
Annual PM Analysis: Details
quarter. The default is 11 days, which corresponds to > 75% completeness for monitors
on a 1 in 6 day schedule. This is a minimum number of samples that is routinely used in
calculations of quarterly average concentrations.
Minimum valid quarters in design value period. This is the number of quarters for which
we have data within three consecutive design value periods. The default value is 12
quarters. If the value is set =11, then there will need to be at least 11 valid quarters (i.e.,
two years must have 4 valid quarters and one year must have at least 3 valid quarters.)
O Official Desiqn Values
® !Qustom.Desjjgn Values:
Valid FRM Quarters
Minimum days for valid quarter
Valid FRM Design Values
Minimum valid quarters in design value period fl2 t
5.6.2.1 Completion Code Use
The Completion Code is a variable in the official quarterly PM2.5 monitor datafile that
MATS uses to identify valid data. As noted in the section on Official Quarterly PM2.5
Monitor Data Input, the completion code has values of: 1, 2, 3, 4, or 5; and the codes are
valid for the end year of each 3 year design value period. MATS uses the Completion
Code variable somewhat differently when using official design values and when using
custom design values.
When using official data, MATS will only use completion codes: 1 and 2.
When using custom data, MATS will potentially use completion codes: 1, 2, 3, and 4 (if
the user specified completion criteria are met).
The following is an explanation of the official EPA completion codes:
Code "1"- complete data and violates the NAAQS
Code "2"- complete data that is below the NAAQS
Code "3"- incomplete data and violates the NAAQS
Code "4" - incomplete data that is below the NAAQS
Code "5" - data that is not comparable to the NAAQS and should not be used. For
example, FRM data collected at a micro-scale site cannot be compared to to the annual
PM2.5 NAAQS.
Abt Associates Inc.
134
June 2010
-------
Annual PM Analysis: Details
5.6.3 Valid FRM Monitors
By default, MATS assumes that there only needs to be one design value for a monitor to be
considered valid. In addition, MATS assumes that no particular design value is required,
so different monitors with different years of data could be used. For example, if you
specify the start year and end year as 2000 and 2004 (giving potential design values of
2002, 2003, and 2004), then one monitor could have data for, say, 2002 and another
monitor data for 2003, and both monitors would be used.
Valid FR.M Monitors
Minimum Number of Design Values
I1 ±f
Required Design Values
| None selected
5.6.4 NH4 Future Calculation
As described in the section on Forecasted Species Calculation. MATS can forecast NH4
using two different approaches. The default approach is to use base year DON values.
NH4 future calculation
(* Calculate future year NH4 using base year (constant) DON values
C Calculate future year MH4 using base year NH4 and the NH4 RRF
5.7 Model Data Options
The Model Data Options section allows you to specify:
Max Distance to Domain. This is the maximum distance from a given monitor to the
nearest model grid cell (measured in kilometers). If a monitor is further than the
specified maximum distance from the center of any grid cell, then MATS will not
calculate results for that monitor or use the monitor in any calculations (this effectively
eliminates monitors from outside the specified domain from the being included in the
output files).
Temporal Adjustment at Monitor. This option specifies how many model grid cells to
use in the calculation of RRFs for point estimates and for spatial estimates. Using the
drop-down menu, you can choose lxl, 3x3, 5,x5, and 7x7. (The default for a 12
kilometer by 12 kilometer grid is to use a 3x3 set of grid cells. The choice of grid size is
discussed in the following EPA guidance:
http://www.epa.gov/scram001/guidance/guide/final-03-pm-rh-guidance.pdn ForPM
analyses, MATS calculates mean concentrations across the grid cell array (as compared
to maximum concentrations used for ozone analyses).
Abt Associates Inc.
135
June 2010
-------
Annual PM Analysis: Details
5.8 Final Check
The Final Check window verifies the selections that you have made.
Abt Associates Inc.
136
June 2010
-------
Annual PM Analysis: Details
Annual PM Analysis
BBS Choose Desired Output
¦ Output Choices - Advanced
¦ Data Input
¦ Species Fractions Options
¦ Species Fractions - Advanced
¦ PM2.5 Calculation Options
¦ Model Data Options
Final Output and Check
Final Output Choices and Verification
Verify inputs
Checking...
Check OK. Press the finish button to continue..
Save Scenario < Back Save Scenario & Run Cancel
Click the button Press here to verify your selections. If there are any errors, MATS will
present a message letting you know. For example, if the path to a model file is invalid —
perhaps you misspelled the file name — you would get the following error:
- Verify inputs
Press here to verify your selections...
Checking...
Please verify the Model Data Baseline File setting.
Check OK. Press the finish button to continue..
After making the necessary correction, click the button Press here to verify your
selections. Then click the Finish button.
Abt Associates Inc.
137
June 2010
-------
Annual PM Analysis: Details
- Verify inputs
[i Press here to.ven^our seiections... i|
Checking...
Check OK. Press the finish button to continue..
5.8.1 Running MATS in Batch Mode
The Save Scenario button will save the scenario as a configuration file (.cfg file). The
"*.cfg" file will be saved in the .\MATS\output directory. Several .cfg files can be created
with the MATS interface and run later in batch mode. To do this, edit the default batch file
located in the .\MATS directory. The file "batchmats.bat" should be edited with a text
editor to point to the name and location of the .cfg files that will be run in batch mode.
& U It raE dit - 3 2 - [C: \Doc ume nts a n d Setti ngs\btimi n\Deskto p\M ATS - 5 -0 8\batc hmats. bat] |"^~| [~] | X J
& File Edit Search Project View Format Column Macro Advanced Window Help _ n1 X
D Or eJ1 S a H A I Wt | H | r X ir=i ® ^mT^IfPgo tp 0 HM
batchmats.bat
start /wait MATS.exe "c:\Program Files\Abt Associates\MATS\\output\default.cfg" ^
start /wait MATS.exe "c:\Prograiri Files\Abt Associates\HATS\\output\default2.cfg"
.A I jf*
For Help, press F1 Ln 1, Col. 1, CW DOS Mod: 5/23/2008 5:02:49PM File Size: 163 INS | ^
After editing the batchmats.bat file, simply run the .bat file. MATS will start and run in
the background.
Abt Associates Inc.
138
June 2010
-------
Daily PM Analysis: Quick Start Tutorial
6 Daily PM Analysis: Quick Start Tutorial
In this tutorial you will forecast daily peak PM2.5 design values at monitors in the Eastern
United States. The steps in this analysis are as follows:
Step 1. Start MATS. Start the MATS program and choose to do an Annual PM
analysis.
Step 2. Output Choice. Choose the output to generate. In this example, you will do two
things: forecast daily peak PM2.5 levels at monitor locations and output a species
fractions file (which you can subsequently reuse, as discussed in the Output Choice
section of the Daily PM Analysis: Details chapter).
Step 3. Output Choice - Advanced. With these advanced options, you can generate a
variety of files useful for quality assurance. Simply review these options and then
uncheck them all. (If you are interested, these options are described in the Output
Choice - Advanced section of the Daily PM Analysis: Details chapter.)
Step 4. Data Input. Choose the particular years of data and monitors to use in this
analysis.
Step 5. Species Fractions Calculation Options. Specify how to generate the relative
response factors fRRFs) used in the forecasts.
Step 6. Species Fractions Calculation Options - Advanced. This window allows you to
make relatively advanced choices for your analysis, such as choosing different ways to
interpolate the monitor data.
Step 7. PM2.5 Calculation Options. Among other things you can specify the particular
years of monitor data that you want to use.
Step 8. Model Data Options. Choose how to use the model data, such as determining
the maximum distance the model data can be from a monitor and identifying peak model
values.
Step 9. Final Check. Verify the choices you have made.
Step 10. Map Output. Prepare maps of your forecasts.
Step 11. View & Export Output. Examine the data in a table format.
Each step is explained below. Additional details are provided in the section Daily PM
Analysis: Details.
6.1 Stepl. Start MATS
Double-click on the MATS icon on your desktop, and the following window will appear:
Abt Associates Inc.
139
June 2010
-------
Daily PM Analysis: Quick Start Tutorial
Umats
Help T
Start Map View Output Navigator
Ozone Analysis
Visibility Analysis
Annual PM Analysis
Daily PM Analysis
Stop Info
Click the Daily PM Analysis button on the main MATS window. This will bring up the
Configuration Management window.
Configuration Management
(* plreate New Configuration!
C Open Existing Configuration
Go
Cancel
A Configuration allows you to keep track of the choices that you make when using MATS.
For example, after generating results in MATS, you can go back, change one of your
choices, rerun your analysis, and then see the impact of this change without having to enter
in all of your other choices. For this example, we will start with a New Configuration.
Choose Create New Configuration and click the Go button. This will bring up the
Choose Desired Output window.
6.2 Step 2. Output Choice
The Choose Desired Output window allows you to choose the output that you would like
Abt Associates Inc.
140
June 2010
-------
Daily PM Analysis: Quick Start Tutorial
to generate. MATS allows you to conduct a Standard Analysis (i.e., forecast Point
Estimates at ambient monitors), output quarterly model data, and output a species fractions
file.
In the Scenario Name box type "Tutorial Daily PM' - this will be used to keep track of
where your results are stored and the variable names used in your results files.
Standard Analysis. Leave the box checked next to "Interpolate monitor data to FRM
monitor sites. Temporally-adjust." MATS will create forecasts for each monitor in the
monitor file.
Quarterly Peak Model Data. Check this option. MATS generates quarterly model files
that MATS generates from daily data that you have provided. This is useful if you want
to reuse model files — the quarterly files are much smaller and MATS will run faster if it
can skip the step of creating quarterly data from the daily.
Species Fraction. Check the box next to Output species fractions file.
Actions on run completion. Check the box next to Automatically extract all selected
output files. Upon completing its calculations, MATS will extract the results into a
folder with the name of your scenario.
Additional details on each of these options are Output Choice section of the Daily PM
Analysis - Advanced chapter.
Daily PM Analysis
Choose Desired Output
¦ 0 utput Choices • Advanced
I Data Input
¦ Species Fractions Options
¦ Species Fractions - Advanced
¦ PM2.5 Calculation Options
¦ Model Data Options
HSfl Final Output and Check
Choose Desired Output
Scenario Name: Tuturial Daily PM|
Standard Analysis
p Interpolate speciation monitor data to FRM monitor sites. Temporally-adjust.
Quarterly Peak Model Data
W Output quarterly peak model data file.
Species Fraction
f Output species fractions file.
Actions on run completion
W Automatically extract all selected output files
< Back
Next >
Cancel
When your window looks like the window above, click Next. This will bring you to the
Abt Associates Inc.
141
June 2010
-------
Daily PM Analysis: Quick Start Tutorial
Output Choice - Advanced window.
6.3 Step 3. Output Choice - Advanced
With the advanced options in the Output Choice - Advanced window, you can generate a
variety of files useful for quality assurance. Simply review these options and then uncheck
them all. (If you are interested, these options are all described in the Output Choice -
Advanced section of the Daily PM Analysis: Details.)
When your window looks like the window above, click Next. This will bring you to the
Data Input window.
6.4 Step 4. Data Input
The Data Input window allows you to choose the species and PM2.5 monitor data and the
model data that you want to use. As discussed in more detail in the following chapter (see
Standard Analysis). MATS calculates the ratio of the base and future year model data to
calculate a relative response factor (RRF) for each PM species. MATS uses the PM2.5
monitor data and interpolated species monitor data to estimate species values at each FRM
site, multiplies the species values from the monitor data with the species-specific RRFs,
and then estimates a future-year design value. (Additional details on Data Input are
available here.)
Use the default settings in the Data Input window. The window should look like the
Abt Associates Inc.
142
June 2010
-------
Daily PM Analysis: Quick Start Tutorial
following:
Daily PM Analysis
BMSm Choose Desired Output
¦ Output Choices - Advanced
Data Input
¦ Species Fractions Options
¦ Species Fractions - Advanced
¦ PM2.5 Calculation Options
¦ Model Data Options
Final Output and Check
Data Input
Species Data
• Species Monitor Data File
fCAProgram FilesSAbt AssociatesSM ATS \S arnpleD ata\Species-f —|
O Species Fractions File [
| spatial field ~ur
~3
PM2.5 Monitor Data
Unofficial Daily Average PM2.5 Data File ffor All Species Fractions & PM2.5 Spatial Fields
| C: \Program Files\Abt Associates\M AT S \S arnpleD ata\PM 25-f or-f ractions-020G-v2. csv
Official Daily Average FRM Data File (for PM2.5 Point Calculations)
| C: \Program Files\Abt Associates\M AT S \S arnpleD ata\24-hr-of f icial-FR M -99-07-v2. csv
Model Data
© Daily model data input O Quarterly peak model data input
Baseline File |C:\Program Files^Abt Associates\M ATSS.SarnpleData\2002cc_EUS_PM25_sub.csv
Forecast File | C: \Program Files\Abt Associates\M AT S\S arnpleD ata\2020cc_E U S_PM 25_sub. csv
< Back
Next >
Cancel
When your window looks like the window above, click Next. This will bring you to the
Species Fractions Calculation Options window.
6.5 Step 5. Species Fractions Calculation Options
The Species Fractions Calculation Options window has several functions related to the
IMPROVE-STN (species) monitor data and the (unofficial) PM2.5 monitor data. These
functions include identifying the years of monitor data that you want to use, deleting any
specific data values, and choosing the minimum data requirements of monitors you want in
your analysis.
Monitor Data Years. Choose the years of monitor data that you want to use. The
default is to use the three-year period 2002-2004. (That is, for both IMPROVE-STN and
PM2.5 monitor data, the Start Year is 2002 and the End Year is 2004.)
Delete Specified Data Values. The default is to delete the observations specified by
EPA. As described in the Data Input section, valid data are given a value of "0" and
observations that should be deleted are given a value of" 1" to "10". (Leave unchecked
the option for the user to flag data.)
Minimum Data Requirements. There are three sets of minimum data requirements:
1. Minimum number of valid days per quarter. This is the minimum number of
Abt Associates Inc.
143
June 2010
-------
Daily PM Analysis: Quick Start Tutorial
site-days per valid quarter. The default is 11 days, which corresponds to > 75%
completeness for monitors on a 1 in 6 day schedule. This is a minimum number
of samples that is routinely used in calculations of quarterly average
concentrations.
2. Minimum number of valid quarters per valid year. This is the minimum
required number of valid quarters for a valid year. The default for
IMPROVE-STN monitor data is 1 quarter, and the default for PM2.5 monitor
data is 4 quarters. If the value is set = 4, then all 4 quarters must be valid in
order for a given year to be considered complete.
3. Minimum number of valid years required for valid monitor. This is the number
of valid years that are needed in order for a particular monitor's data to be
considered valid. The default is 1 for both IMPROVE-STN and PM2.5 monitor
data. For example, if the value is = 1, then a monitor's data will be used in the
species fractions calculations if it has at least one valid year. If the value = 3,
then the monitor must have 3 years of valid data to be used.
Use the default settings pictured in the screenshot below. (All of these options are
described in detail here.')
Abt Associates Inc.
144
June 2010
-------
Daily PM Analysis: Quick Start Tutorial
Daily PM Analysis
Choose Desired Output
¦ Output Choices - Advanced
¦ Data Input
Species Fractions Options
¦ Species Fractions - Advanced
¦ PM2.5 Calculation Options
¦ Model Data Options
El Final Output and Check
Species Fractions Calculation Options
IMPROVE-STN Monitor Data
Monitor Data Years
Start Year End Year
12002 ~~E] 12004
Delete Specified Data Values
|v EPA-specified deletions from monitor data
User-specified deletions from monitor data
Minimum Data Requirements
Minimum number of valid days per quarter
Minimum number of valid quarters per valid year
Minimum number of valid years required for valid monitor
PM2.5 Monitor Data
Monitor Data Years
Start Year End Year
12002 _d ^004 3
Delete Specified Data Values
[~ EPA-specified deletions from monitor data
User-specified deletions from monitor data
Minimum Data Requirements
Minimum number of valid days per quarter
Minimum number of valid quarters per valid year (point calculations) J~~
Minimum number of valid years required for valid monitor
~ :
«a
^±i
< Back
Next >
Cancel
When your window looks like the window above, click Next. This will bring you to the
Species Fractions Calculation Options - Advanced window.
6.6 Step 6. Species Fractions Calculation Options -
Advanced
The Species Fractions Calculation Options - Advanced screen allows you to make
relatively advanced choices for your analysis. Generally speaking, the default options
settings are consistent with the EPA modeling guidance document. The first set of options
lets you specify which monitor data you want to use to characterize peak values (e.g., top
10 percent of daily monitor days). The second set of options allows you to specify the
interpolation weighting that you want to use and whether the interpolation involves a
maximum distance or not. The third set of options involves choices regarding ammonium,
blank mass, and organic carbon.
Use the default settings pictured in the screenshot below. (All of these options are
described in the Species Fractions Calculation Options - Advanced section in the Daily PM
Analysis: Details chapter.)
Abt Associates Inc.
145
June 2010
-------
Daily PM Analysis: Quick Start Tutorial
Daily PM Analysis
Choose Desired Output
¦ Output Choices - Advanced
¦ Data Input
¦ Species Fractions Options
Species Fractions - Adv
¦ PM2.5 Calculation Options
Bfl Model Data Options
Fmal Output and Check
Species Fractions Calculation Options - Advanced
11
Using Monitor Data to Calculate Species Fractions
IMPROVE-STN Monitor Data
® jUse top X percent of daily monitor days;
Use all daily monitor values greater than
fixed amount (uq/m3)
Minimum number of days required above fixed amount
Use top X number of daily monitor days
PM2.5 Monitor Data
~ Use top X percent of daily monitor days
Use all daily monitor values greater than
fixed amount (uq/m3)
Minimum number of days required above fixed amount
Use top X number of daily monitor days
Interpolation Options
i?e Distance Squared V " I |90QQQ "l~H '•
10-M
25-r|
log
~°±l
1±l
isg
PM2.5
S04
N03
EC
Salt
3 D istance S quared 190000 »
Inverse Distance Squared 190000
Inverse Distance Squared V ~ 190000 »
Inverse Distance Squared V 190000
Miscellaneous Options
Ammonium
• Use DON values
O Use measured ammonium
NH4 percentage evaporating (0-100)
Crustal j Inverse Distance Squared vT] |90000 «|^»]
DON | Inverse Distance Squared V ~ [ 190000 «|-^-| »(
OC | Inverse Distance Squared V * | j90000 ¦« EH
NH4 [inverse Distance Squared v [90000 "j-rj »|
r
*3
Default Blank Mass
Default Blank Mass
Organic Carbon
Organic carbon mass balance floor [
Organic carbon mass balance ceiling | 0.8~^~|
0.5-^-j
1±l
When your window looks like the window above, click Next. This will bring you to the
PM2.5 Calculation Options window.
6.7 Step 7. PM2.5 Calculation Options
The PM2.5 Calculation Options window allows you to specify the particular years of
monitor data that you want to use from the input file you specified in Step 4 (Data Input)
Keep the default settings:
PM2.5 Monitor Data Years. Leave Start Year = 2000 and End Year = 2004.
Abt Associates Inc.
146
June 2010
-------
Daily PM Analysis: Quick Start Tutorial
Valid FRM Monitors. Keep the minimum number of design values equal to the default
value of 1, and do not specify any particular design values for inclusion in the
calculations.
NH4 Future Calculation. You can also specify how you want to forecast NH4 values.
Use the default approach, which is to use baseline DON values.
Use the default settings pictured in the screenshot below. (All of these options are
described in detail here.)
When your window looks like the window above, click Next. This will bring you to the
Model Data Options window.
6.8 Step 8. Model Data Options
The Model Data Options section allows you to specify three items:
Temporal Adjustment at Monitor. This option specifies how many model grid cells to
use in the calculation of RRFs for point estimates. Use the default option: lxl set of grid
cells. Note that for PM analyses, MATS calculates mean concentrations across the grid
cell array. (Compare this with the maximum concentrations used for ozone analyses,
however, in the case of a lxl set of grid cells the mean and maximum are the same.)
Advanced Options: RRF Model Values Used. This option lets you specify which
monitor data you want to use to characterize peak values (e.g., top 10 percent of daily
model days)
Abt Associates Inc.
147
June 2010
-------
Daily PM Analysis: Quick Start Tutorial
Max Distance to Domain. This is the maximum distance from a given monitor to the
nearest model grid cell (measured in kilometers). If a monitor is further than the
specified maximum distance from the center of any grid cell, then MATS will not
calculate results for that monitor or use the monitor in any calculations (this effectively
eliminates monitors from outside the specified domain from the being included in the
output files).
Use the default settings pictured in the screenshot below. (All of these options are
described further here.)
Daily PM Analysis
Choose Desired Output
Output Choices - Advanced
Data Input
Species Fractions Options
Species Fractions - Advanced
Model Data Options
Final Output and Check
Model Data Options
Temporal adjustment at monitor
Grid for Point Forecast Statistic
pTxi 3 lMeari 3
Advanced Options: RRF Model Values Used
RRF-model values used
® Use top X percent of daily model days
Use all daily model values greaterthan
fixed amount (uq/m3)
Minimum number of days required above fixed amount [
O Use top X number of daily model days
1(H
~°5
"^E
Man Distance to Domain [km] |~
25
< Back
Next >
Cancel
When your window looks like the window above, click Next. This will bring you to the
Final Check window.
6.9 Step 9. Final Check
The Final Check window verifies the choices that you have made. For example, it makes
sure that the paths specified to each of the files used in your Configuration are valid.
Click on the Press here to verify selections button.
Abt Associates Inc.
148
June 2010
-------
Daily PM Analysis: Quick Start Tutorial
Daily PM Analysis
Choose Desired Output
¦ Output Choices-Advanced
¦ Data Input
¦ Species Fractions Options
¦ Species Fractions - Advanced
¦ PM 2.5 Calculation 0 ptions
¦ Model Data Options
Final Output and Check
D
Final Output Choices and Verification
Verify inputs
| Press here
Checking...
Check OK. Press the finish button to continue..
Save Scenario < Back Save Scenario & Run
If you encounter any errors, go back to the choices you have previously made by clicking
on the appropriate part (e.g., Data Input) of the tree in the left panel, and then make any
changes required.
When your window looks like the window above, click either Save Scenario & Run or
Save Scenario. Save Scenario & Run will cause MATS to immediately run the scenario.
A temporary, new Running tab will appear (in addition to the Start, Map View and
Output Navigator tabs).
Abt Associates Inc.
149
June 2010
-------
Daily PM Analysis: Quick Start Tutorial
MATS
EES
Help "
I Start Map View Output Navigator
Running |
I
Close i|
Name
| Last Message
Tutorial Daily PM.asr
Loading/creating model baseline data.
When the calculations are complete, a small window indicating the results are Done will
appear. Click OK.
Done
Mill
After clicking OK, MATS will open a folder with the results files already exported.
Abt Associates Inc.
150
June 2010
-------
Daily PM Analysis: Quick Start Tutorial
C:\Prograrn FilesUbt Associates\MATS\output\Tutorial Daily PM
File Edit View Favorites
Tools Help
Qeack • Q ¦
Search Folders
[03'
Address 1^) C:\Prograrn Files\Abt Associates\MATS\output\Tutorial Daily PM
File and Folder Tasks
Make a new folder
fcl Publish this folder to the
Web
© Share this folder
Other Places
A j
°utPut
Q My Documents
PHi Shared Documents
My Computer
My Network Places
Name
B Configuration .cfg
||] Log File,log
Tutorial Daily PM Baseline Quarterly Peak Model Data.csv
Tutorial Daily PM Daily PM25 Point.csv
Tutorial Daily PM Future Quarterly Peak Model Data.csv
Tutorial Daily PM Used Baseline Quarterly Peak Model Data Point.csv
Tutorial Daily PM Used Future Quarterly Peak Model Data Point.csv
Size Type
224
KB
CFG File
2
KB
Text Document
,114
KB
Microsoft Excel
Comma
Separat.
46
KB
Microsoft Excel
Comma
Separat.
,107
KB
Microsoft Excel
Comma
Separat.
115
KB
Microsoft Excel
Comma
Separat.
115
KB
Microsoft Excel
Comma
Separat,
Output Navigator tab will also be active. (The Running tab will no longer be seen.)
MATS will automatically load the output files associated with the .asr configuration that
just finished running.
g MATS
Help
0®®
Start Map View | Output Navigator I
Load
Extract All
Highlight file of interest and right-click to view options to Map, View, and Extract the data.
I Type
Name
- Configuration/Log Files
h Configuration Configuration
: Log File Run Log
- Output Files
Tutorial Daily PM Baseline Quarterly Peak Moc Monitor Network
Tutorial Daily PM Future Quarterly Peak Model Monitor Network
Tutorial Daily PM Used Baseline Quarterly Pea Monitor Network
Tutorial Daily PM Used Future Quarterly Peak N Monitor Network
Tutorial Daily PM Daily PM25 Point Monitor Network
223kb
1 kb
2113kb
2106kb
1"Mkb
1"Mkb
45kb
Stop Info
Abt Associates Inc.
151
June 2010
-------
Daily PM Analysis: Quick Start Tutorial
The next step (click here') shows you how to map your results with the Output Navigator.
For more details on mapping and other aspects of the Output Navigator, there is a
separate chapter on the Output Navigator.
6.10 Step 10. Map Output
After generating your results, Output Navigator can be used to load and/or map them. If a
run just finished, the output files will already be loaded into output navigator. (If files from
a previous run need to be loaded then click on the Load button and choose the Tutorial
Daily PM.asr file.)
Under Configuration/Log Files, you will see two files:
Configuration, keeps track of the assumptions that you have made in your analysis.
Los File : provides information on a variety of technical aspects regarding how a results
file (*.ASR) was created.
Under Output Files you will see four files with model data and one with forecasted design
values:
Tutorial Daily PM Baseline Quarterly Peak Model Data: contains baseline peak
quarterly species and PM2.5 values for all grid-cells in the modeling domain.
Tutorial Daily PM Future Quarterly Peak Model Data: contains future peak quarterly
species and PM2.5 values for all grid-cells in the modeling domain.
Tutorial Daily PM Used Baseline Quarterly Peak Model Data Point: contains only
baseline peak quarterly species and PM2.5 values that were used in the analysis.
Tutorial Daily PM Used Future Quarterly Peak Model Data Point: contains only
baseline peak quarterly species and PM2.5 values that were used in the analysis.
Tutorial Daily PM Daily PM25 Point: contains peak base & future PM2.5, species
values, and RRFs. (Note that the annual RRFs and annual species values are not used
anywhere in the calculation of design values, and here just for information.)
Right-click on the file Tutorial Daily PM Daily PM25 Point. This gives you three
options: Add to Map, View, and Extract. Choose the Add to Map option.
Abt Associates Inc.
152
June 2010
-------
Daily PM Analysis: Quick Start Tutorial
Help
Start Map View | Output Navigator I
Load
Extract All
Highlight file of interest and right-click to view options to Map, View, and Extract the data.
[Name
I Type
Size
B Configuration/Log Files
1 h Configuration Configuration
Log File Run Log
- Output Files
Tutorial Daily PM Baseline Quarterly Peak Model Data Monitor Network
Tutorial Daily PM Future Quarterly Peak Model Data Monitor Network
Tutorial Daily PM Used Baseline Quarterly Peak Model Data Point Monitor Network
Tutorial Daily PM Used Future Quarterly Peak Model Data Point Monitor Network
223kb
1 kb
2113kb
2106kb
114kb
114kb
Tutorial Daily PM Daily
Monitor Network
View
Extract
This will bring up the Map View tab.
To view an enlarged map, use the Zoom to an area Task Bar button on the far left.
Abt Associates Inc.
153
June 2010
-------
Daily PM Analysis: Quick Start Tutorial
Choose the Continental US.
Help
Start | Map View | Output Navigator
+ -"V-v"1
Full Extent
Maryland
New England
Southern California
Texas
Washington DC
Edit Zoom Frames
Add Current View to List
Long:-193.71675, Lat: 34.09601 ***
Extent: Min(-159.661,2.334) Max(-**W51.197)
stop Info
To more easily view the location of monitors in particular states, uncheck US Counties
using the Standard Layers drop down menu on the far right of the Task Bar. Your
window should look like the following:
Abt Associates Inc.
154
June 2010
-------
Daily PM Analysis: Quick Start Tutorial
~
Start | Map View | Output Navigator
+ ' i"'lj ' j^j ^ Standard Layers "
JjData Loaded j|
@ • Tutorial Daily PM Daily PM25
Long: -113.26993, Lat: 52.11875 ***
Extent: Min(-122.607,21.608) Max»**7.820,45.788)
Zoom in further on the Eastern US using the Zoom in button on the Task Bar. This allows
you to view the results more closely. A dashed line surrounds the area that you have
chosen and should look something like the following:
Abt Associates Inc.
155
June 2010
-------
Daily PM Analysis: Quick Start Tutorial
V MATS
Start | Map View | Output Navigator
K N' g"|j [3 standard Layers-
jDSaLoadedj|
0 ~ Tutorial Daily PM Daily PM25 F
Long: -76.84206, Lat: 29.07562 ^
Extent: Min(-122.607,21.608) Max***7.820,45.788)
Right click on the "Tutorial Annual PM Annual PM25 Point" layer in the panel on the left
side of the window. Choose the Plot Value option.
ifMATS BEE
Help "
Start | Map View | Output Navigator
+SJ +-v 1 : " ^ Standard Layers -
Long: -94.34759, Lat: 35.83098
Extent: Min(-95.053,31.161) Max(-***.249,38.263)
Stop Info
Abt Associates Inc.
156
June 2010
-------
Daily PM Analysis: Quick Start Tutorial
This will bring up Shape Class Breaks window. In the Value drop-down list, choose the
variable "f_pm25 aim dv" - this is forecasted PM2.5 design value. Set the Marker
Sizing to 1 - this will give larger values a somewhat larger marker on the map.
Shape Class Breaks
Layer Name:
Tutoriai Daily PM Daily PM25 Point
Value:
f_prn25_d_dv
J
Date
no date
J
Bins
r Unique Values
Class Count:
|5 1]
Marker Sizing: 1
I ^
Start Color
End Color
iJS Clear Breaks V' Apply X Close
Click Apply and then click Close. This will bring you back to the Map View window.
BUM
Help ^
Start | Map View | Output Navigator
Long:-92.00597, Lat: 42.40581 ***
Extent: Min(-95.053,31.161) Max(^**249,38.263)
+. " +, ' |||| Standard Layers "*¦
J; Data Loaded ;|
@ • Tutoriai Daily PM Daily PM2u r
© • f_pm25_d_dv 16.5 to 22.2
@ • f_pm25_d_dv 22.2 to 24.2
0 © f_pm25_d_dv 24.3 to 26.4
@ O f_pm25_d_dv 26.5 to 28.5
@ O f_pm25_d_dv 29 to 36.3
This is just a brief summaiy of the mapping possibiliti es available. For more details, there
is a separate chapter on the Map View. The next step is to go to the Output Navigator to
Abt Associates Inc.
157
June 2010
-------
Daily PM Analysis: Quick Start Tutorial
view the data in a table format.
6.11 Step 11. View Output
After mapping your results, click on the Output Navigator tab, so that you can then view
the data in a table. Right-click on the file Tutorial Daily PM Daily PM25 Point. This
gives you three options: Add to Map, View, and Extract. Choose the View option.
Help
Start Map View I Output Navigator I
Load
Extract All
Highlight file of interest and right-click to view options to Map, View, and Extract the data.
[Name
|Type
| Size
0 Configuration/Log Files
j- Configuration
Configuration
223kb
! Log File
Run Log
1 kb
[=1 Output Files
Tutorial Daily PM Baseline Quarterly Peak Model Data
Monitor Network
2113kb
Tutorial Daily PM Future Quarterly Peak Model Data
Monitor Network
2106kb
Tutorial Daily PM Used Baseline Quarterly Peak Model Data Point Monitor Network
1Hkb
Tutorial Daily PM Used Future Quarterly Peak Model Data Point
Monitor Network
TMkb
Monitor Network
45kb
Add 1 o map
Extract
Stop Info
This will bring up a Monitor Network Data tab. The upper left panel allows you to view
the ID and latitude and longitude of the monitors in your data — at the right of this panel
there is a scrollbar with which you can locate any particular monitor of interest.
Abt Associates Inc.
158
June 2010
-------
Daily PM Analysis: Quick Start Tutorial
Help ^
Start Map View Output Navigator [ Monitor Network Data |
Tutorial Daily PM Daily PM25 Point
Show All or select a particular location to see data.
id
010270001
010331002
010491003
010550010
"Tvi
pe
010730023
010731005
010731009
m nTJonn?
FRM
FRM
FRM
FRM
FRM
FRM
FRM
¦^C>M
33.281261
34.760556
34.287627
33.993749
[long
-85.802182
-87.650556
-85.968298
-85.991072
33.553056
33.331111 j
33.459722
-86.815
-87.305556 r
017 PO.11 CI
S
Select Quantities that must be >= 0
monitor_gridcell
b_pm25_d_dv
f_pm25_d_dv
b_blank_mass
b_crustal_mass
b_ec_mass
b_nh4_mass
b_ocmb_mass
b_so4_rnass
b_no3_mass
b_water_mass
b_salt_mass
f blank mass
_H|
Export Exportcurrentlyshown datato CSV
d ate | m o n ito r_g ri (| b_p m 2 5_d_ | f_p m 2 5_d_c | b_b I an k_rm | b_cru stal_rr j b_e c_m as s | b_n h4_mas|b_ocm b_m ¦ | b_s o 4_m as | b_n o 3_m as | b
id
010270001
010331002
010491003
010550010
no date
no date
no date
no date
171079
15509lj
169088
169085
30.3 j
31.5]
33.71
36.01
23.2
25.5
25.6
27.5
010730023
010731005
-------
Daily PM Analysis: Quick Start Tutorial
Help "
Start Map View 1 Output Navigator | Monitor Network Data |
Tutorial Daily PM Daily PM25 Point
Close
Show All | or select a particular location to see c
1 |VPe
[long
010270001
FRM
33.281261
-85.802182
010331002
FRM
34.760556
-87.650556
010491003
FRM
34.287627
-85.968298
010550010
FRM
33.993749
-85.991072
010730023
FRM
33.553056
-86.815
010731005
FRM
33.331 111
-87.003611
010731009
FRM
33.459722
-87.305556
nimT?nm I
cnu
tj rinniaal
Select Quantities that must be >= 0
0L
monitor_gridcell
b_pm25_d_dv
f_pm25_d_dv
b_blank_mass
b_crustal_mass
b_ec_mass
b_nh4_mass
b_ocmb_mass
b_so4_mass
b_no3_mass
b_water_mass
b_salt_mass
f blank mass
M
Export Exp o rt cu rre ntly s h own d ata to CSV
date
| m o n ito r_g ri (| b_p m 2 5_d_ | f_p m 2 5_d _c | b_b I an k_m q b_cru staljr | b_e c_m as s | b_n h-!l_mas|b_ocm b_m i | b_s o 4_m as | b_n o 3_m as | b_we
010491003
no date
169088
33.7
25.6
0.500
1.512
1.306
3.081
10.71
7.173
1.87
IT
Stop Info
To view all of the data again, click on the Show All button.
For additional details on generating annual PM results, see the chapter on Daily PM
Analysis: Details. For additional details on viewing data, see the View Data section in
chapter on the Output Navigator.
Abt Associates Inc.
160
June 2010
-------
Daily PM Analysis: Details
7 Daily PM Analysis: Details
MATS can forecast daily design values at PM2.5 monitor locations — these forecasts are
referred to as Point Estimates. The set of choices involved in calculating Point Estimates
can be fairly involved, so MATS keeps track of these choices using a Configuration.
When you begin the process of generating PM2.5 estimates, MATS provides an option to
start a new Configuration or to open an existing Configuration.
Configuration Management
(* ICreate New Configuration!
r Open Existing Configuration
Go
Cancel
Select your option and then click Go.
MATS will then step you through a series of windows with choices for your analysis.
Output Choice. Choose whether you want to run the Standard Analysis, and whether to
output a species fractions file and/or quarterly model data.
Output Choice - Advanced. This option provides miscellaneous Point Estimate output
files that are mainly used for QA.
Data Input. Load species monitoring data or a species fractions file. LoadPM2.5
ambient monitoring data. And, load the modeling data that you want to use.
Species Fractions Calculation Options. Choose the years of daily STN-IMPROVE and
FRM monitoring data. Identify valid monitors. Delete specified values.
Species Fractions Calculation Options - Advanced. Choose method to identify peak
monitor values. Choose interpolation options for PM2.5 and species monitoring data.
Choose assumptions for the ammonium calculation, default blank mass, and organic
carbon.
PM2.5 Calculation Options - FRM Monitor Data. Choose the years of quarterly FRM
monitoring data. Identify valid monitors. Choose the approach for calculating future
NH4.
Model Data Options. Specify the maximum distance of monitors from modeling
domain. Choose method to identify peak model values. Specify which model grid cells
will be used when calculating RRFs at monitor locations.
Final Check. Verify the selections that you have made.
7.1 Output Choice
In the Output Choice window, MATS lets you specify the name of your Scenario, and
Abt Associates Inc.
161
June 2010
-------
Daily PM Analysis: Details
then to choose up to three file options: Standard Analysis, which refer to forecasts made at
FRM PM2.5 monitor locations; Quarterly Model Data, which allows you to create
quarterly averages from daily model output data (output data from grid models such as
CMAQ and CAMx), and then subsequently reuse this file; and Species Fractions, which
outputs a reusable species fractions file.
By checking the box next to Automatically extract all selected output files, MATS will
create a separate folder with your chosen Scenario Name in the MATS "Output" folder,
and then export .CSV files with the results of your analysis. Alternatively, you can export
the results from the Output Navigator, but checking this box is a little easier.
Choose Desired Output
Scenario Name: |Exarnple Daily PM
Standard Analysis
|7 Interpolate speciation monitor data to FRM monitor sites. Ternporally-adjust.
Quarterly Peak Model Data
Output quarterly peak model data file.
Species Fraction
W ^Output species fractions file
Actions on run completion
Automatically extract all selected output files
Daily PM Analysis
Choose Desired Output
¦ ~ utput Choices - Advanced
¦ Data Input
¦ Species Fractions Options
¦ Species Fractions - Advanced
¦ PM2.5 Calculation Options
¦ Model Data Options
Final Output and Check
Standard Analysis. The Standard Analysis refers to the calculation of future year
PM2.5 design values at FRM monitor locations. This is the main part of the modeled
attainment test for PM2.5. There are several calculations involved in this analysis. MATS
will interpolate PM2.5 species data, calculate species concentrations at each FRM site and
project design values to a future year using gridded model data. Most MATS users will
run this analysis and it is therefore checked by default.
Quarterly Model Data. MATS requires two types of data input: ambient monitor
data and gridded model output data. For the daily PM2.5 calculations, MATS will accept
either MATS formatted daily average gridded model files or quarterly average files. If
daily average model files are used as inputs, MATS will calculate quarterly averages from
the daily averages and optionally output the quarterly average concentrations into text files
(CSV files). The quarterly average text files can then be re-used in subsequent MATS
analyses. Quarterly average input files are smaller and run faster than daily average files.
Abt Associates Inc.
162
June 2010
-------
Daily PM Analysis: Details
Choosing this option will have MATS generate two types of quarterly average model
concentration CSV files:
All Data. MATS outputs quarterly peak data for all grid cells in the modeling domain.
MATS will create one baseline year file and one future year file. This will create
relatively large files, but they will still be much smaller than daily average files.
Used Data. MATS outputs quarterly peak data for the grid cells that are subsequently
used in the particular MATS configuration. For example, if MATS calculates future
year design values at 20 FRM sites using a 1 X 1 grid array, then MATS will output base
and future model values for only 20 grid cells (assuming each monitor is in a unique grid
cell). The advantage of these files is that they are extremely small. But if subsequent
MATS runs use a different set of monitors or grid arrays, then the files may not contain
all of the necessary data to complete the analysis. This option is recommended as a QA
tool to examine the grid cells and the model concentrations that MATS is using in the
analysis.
Species Fraction. Checking the "Output species fraction file" box will create an output
file containing the calculated PM2.5 species fractions at each FRM site used by MATS.
This species fraction file can be re-used in MATS as an input file. The species fraction file
can be useful for several reasons. One, using a species fraction file saves time because
MATS won't have to interpolate species data and calculate fractions each time it is run.
Two, it can provide consistency between MATS runs by ensuring that the same species
fractions are used each time. And for the same reason, the species fraction file can be used
interchangeably between different users to ensure that multiple groups are using the same
species fractions (if that is a goal). And finally, the fractions file can serve as a template
for creating a custom species fractions file using whatever data and techniques (e.g.
alternative interpolation techniques) are desired by any particular user.
7.1.1 Scenario Name
The Scenario Name allows you to uniquely identify each analysis that you conduct. It is
used in several ways.
Results file name. The results file is given the Scenario Name (e.g., Example Daily
PM.asr). Note that the extension f .ASR) is specifically designated just for MATS and
can only be used by MATS.
Organize output. In the Output folder, MATS will generate a folder using the Scenario
Name. MATS will use this folder as a default location for files generated with this
Scenario Name.
Abt Associates Inc.
163
June 2010
-------
Daily PM Analysis: Details
C:\Program FilesYAbt Associates\MATS\output\Example Daily PM
File Edit View Favorites Tools Help
^ l
O ¦ (5" / Search
Folders
ma-
| Address C:\Program Files\Abt Associates\MATS\output\Example Daily PM
v 0 Go
I Folders x
Name
Size Type
S-Q MATS
configs
data
O help
jfi) maps
0 l£| output
Q Example Annual PM
Q Example Daily PM
H& Example 03
\r^i Example Visibility
|£) Tutorial Annual PM
Q Tutorial Daily PM
Tutorial 03
O Tutorial Visibility
GB Ql sampledata
. i.efg 223 KB CFG File
^Example Daily PM Baseline Quarterly Peak Model Data.csv 2,114 KB Microsoft Excel Comma Si
^Example Daily PM Daily All Years All Quarters PM25 Point.csv 936 KB Microsoft Excel Comma Si
^Example Daily PM Daily All Years High Quarters PM25 Point.csv 80 KB Microsoft Excel Comma Si
^Example Daily PM Daily PM25 Point.csv 46 KB Microsoft Excel Comma Se
^Example Daily PM Future Quarterly Peak Model Data.csv 2,107 KB Microsoft Excel Comma Se
^Example Daily PM High County Sites.csv 35 KB Microsoft Excel Comma S«
^Example Daily PM Neighbor File Point.csv 34,099 KB Microsoft Excel Comma Si
^Example Daily PM Quarterly Peak NH4_DON Monitors.csv 48 KB Microsoft Excel Comma Si
^Example Daily PM Quarterly Peak Spec Frac Point.csv 974 KB Microsoft Excel Comma Si
^Example Daily PM Quarterly Peak Speciated Monitors.csv 137 KB Microsoft Excel Comma St.
^Example Daily PM Used Baseline Quarterly Peak Model Data Point.csv 115 KB Microsoft Excel Comma Si
^Example Daily PM Used Future Quarterly Peak Model Data Point.csv 115 KB Microsoft Excel Comma Si
||3 Log File, log 2KB Text Document
>
Output file names. The output files generated will begin with the Scenario Name.
V MATS
Help
~El®
Start Map View | Output Navigator I
Highlight file of interest and right-click to view options to Map, View, and Extract the data.
| Name
|Type
Size
m
Configuration/Log Files
Configuration
Configuration
222kb
Log File
Run Log
lkb
- Output Files
Example Daily PM Baseline Quarterly Peak Model Data
Monitor Network
2113kb
Example Daily PM Future Quarterly Peak Model Data
Monitor Network
2106kb
Example Daily PM Quarterly Peak Speciated Monitors
Monitor Network
136kb
Example Daily PM Quarterly Peak NhM/DON Monitors
Monitor Network
51 kb
Example Daily PM Used Baseline Quarterly Peak Model Data Point Monitor Network
114kb
Example Daily PM Used Future Quarterly Peak Model Data Point
Monitor Network
114kb
Example Daily PM Quarterly Peak Spec Frac Point
Monitor Network
973kb
Example Daily PM Neighbor File Point
Monitor Network
34991kb
Example Daily PM Daily All Years All Quarters PM25 Point
Monitor Network
935kb
Example Daily PM Daily All Years High Quarters PM25 Point
Monitor Network
79kb
Example Daily PM Daily PM25 Point
Monitor Network
Ahkb
Example Daily PM High County Sites
Monitor Network
34kb
_J Stop Info
7.1.2 Standard Analysis
Future-year PM2.5 design values are calculated at each FRM monitoring site through a
series of calculations:
Abt Associates Inc.
164
June 2010
-------
Daily PM Analysis: Details
Step 1. Baseline Quarterly Peak PM2.5 Calculation. MATS uses "official" daily PM2.5
data to calculate a peak PM2.5 value for each quarter.
Step 2. Baseline Quarterly Peak Species Calculation. MATS calculates species fractions
using "unofficial" PM2.5 monitor data and speciated monitor data to calculate species
fractions. It then multiplies the species fractions with the baseline PM2.5 concentration
calculated in Step 1.
Step 3. Forecasted Quarterly Peak Species Calculation. Multiplying relative response
factors (RRFs) with baseline quarterly species values, MATS calculates forecasted
quarterly peak species values at each FRM monitor site.
Step 4. Forecasted Design Value Calculation. Using the forecasted quarterly peak species
values, MATS calculates the forecasted design value at each FRM monitor site.
In this section, we go into some detail describing these steps. However, you should note
that MATS gives you a number of options affecting the exact steps that MATS follows,
such as determining which years of monitoring data to use and to choosing which monitors
to include in the calculations. These options are detailed in the PM2.5 Calculation Options
section. The output from the Standard Analysis is described here.
7.1.2.1 Step 1: Baseline Quarterly Peak PM2.5 Calculation
The first step in projecting the daily design value is to identify the maximum daily average
PM2.5 concentration in each quarter that is less than or equal to the annual 98th percentile
value over the entire year, using the ("official") daily average PM2.5 file (described in the
Data Input section). This results in data for each year (2000-2004 by default) for each
FRM site which contains one quarter with the 98th percentile value and three quarters with
maximum values less than or equal to the 98th percentile value.
7.1.2.2 Step 2: Baseline Quarterly Peak Species Calculation
Since the FRM monitors do not have speciated data (and the majority of FRM sites are not
co4ocated with a speciation monitor), MATS uses speciated PM2.5 monitor data from
other monitoring networks, such as STN and IMPROVE, to estimate the PM2.5
attributable to the following species: sulfate (S04), nitrate (N03), elemental carbon (EC),
organic carbon (OC), crustal matter, particle bound water (PBW), ammonium (NH4), and,
data permitting, salt.
To estimate these species, MATS uses the "SANDWICH" process (Frank, 2006)
(SANDWICH stands for Sulfates, Adjusted Nitrates, Derived Water, Inferred
Carbonaceous mass, and estimated aerosol acidity [H+]). The default species input data
file contains aerosol nitrate data (N03r) that has been adjusted to account for
volatilization. Additional SANDWICH adjustments are made within MATS. These
include calculation of particle bound water (PBW) and organic carbon by mass balance
(OCMmb).
Determine Valid Days and Calculate Peak Averages for Speciated Monitor Data at
Abt Associates Inc.
165
June 2010
-------
Daily PM Analysis: Details
Speciated Monitors
A first step in calculating baseline quarterly peak species values is deleting invalid days of
data. MATS deletes days from the speciated monitor data that are missing
"MEASUREDFM" values. Days that have a "EPA FLAG" value greater than "0" are
also deleted. MATS then uses the Minimum Data Requirements specified on the Species
Fractions Calculations Options tab, to delete data that are incomplete for a particular
species. In the default case, there should be 11 days each quarter, at least 1 valid quarter
per valid year, and at least 1 valid year per valid monitor. In the default case, only the
specification of a minimum number of values per quarter has any effect on the data. If a
particular quarter in a given year for a given species does not have 11 non-missing species
values, then data for that species is dropped for that particular quarter at that monitor.
The number of peak days is then calculated. In the default case, the top 10% of days =
round(0.10*non-missing days in quarter). The speciated monitor data is then sorted on
"MEASURED FM" at each monitor, for each year and quarter, and the peak days are
chosen. (Options for choosing peak days are described in the Species Fractions
Calculation Options - Advanced section.)
MATS then averages the peak daily values in each year-quarter, and then averages the
values across years. For example, if for 2002-2004 we have S04 averages of peak values
in quarter 1 of 8.2 ug/m3 (for 2002), 6.2 ug/m3 (for 2003), and 3.6 ug/m3 (for 2004), then
the average across the years would equal 6.0 (=[8.2+6.2+3.6]/3).
Interpolate Peak Quarterly Species Values to FRM Monitors
About 75 percent of FRM monitors are not co-located with an STN monitor, so the
estimation of the quarterly peaks of the individual species at these FRM sites depends on
the interpolated quarterly peaks from speciated monitors (e.g., IMPROVE). Individual
species are interpolated to the latitude and longitude associated with each FRM monitor.
(For FRM monitors that are co-located with an STN monitor, MATS simply uses the
species values from the co-located STN monitor.)
For the speciated monitors that are not co-located, MATS uses the Voronoi Neighbor
Averaging (VNA) procedure to interpolate the peak quarterly species values to FRM
monitors. The VNA procedure identifies the nearby ("neighbor") speciated monitors for
each FRM site, and then assigns a weight to each speciated monitor that gets used to
generate an initial, weighted-average for each species at each FRM site. Generally
speaking, speciated monitors that are relatively far away get a smaller weight. (You can
find details on the interpolation process and different options for interpolation at the
section on Interpolation Options for Species Fractions Calculation.)
After Interpolation to FRM Monitors: SANDWICH and Species Fractions
Calculations
After the interpolation of quarterly peak values for retained N03 (N03r), S04, OCM,
crustal, EC, and DON, a few additional steps are necessary to generate speciated quarterly
peak values at each FRM monitor site. These include calculating retained NH4 (NH4r),
Abt Associates Inc.
166
June 2010
-------
Daily PM Analysis: Details
PBW, blank mass, and organic carbon mass (OCMMB), the latter of which is calculated
through a mass balance approach. (The following sections describe the calculations of
ammonium. PBW. and organic carbon [by difference], as well as the blank mass
assumption.)
After generating quarterly peak values at each FRM monitor site, MATS can then calculate
the species fractions. (The species fractions calculations are straightforward and are
described further in the Species Fractions section.)
Summary of Calculations After Interpolation of STN & IMPROVE Speciated Monitor Data
Calculation Description
Calculate Retained Ammonium Calculate ammonium associated with retained nitrate (N03r) and
(NH4r) S04. MATS calculates NH4r using DON, S04, and N03r.
(Alternatively MATS can use directed measured ammonium).
Calculate Particle Bound Water Calculate amount of water associated with ammonium sulfate and
(PBW) ammonium nitrate, which are hygroscopic.
Estimate Blank Mass Account for contamination in FRM monitors.
Calculate Organic Carbon Mass Calculate organic carbon mass with a mass balance approach.
(OCMMB)
Calculate Species Fractions Divide species estimates for S04, N03r, OCMMB, EC, crustal
material, NH4r, and PBW by the non-blank PM2.5 mass. (The
inclusion of salt is optional and is not included in the default MATS
data.)
Using Species Fractions to Estimate Baseline Quarterly Peak Species Values
The use of species fractions to calculate species concentrations is straightforward. The
weighted quarterly species average is calculated by multiplying quarterly peak FRM
baseline values (calculated in Step 1, minus the assumed blank mass, specified by the user)
with species fractions that have been estimated for each FRM monitor. The calculation is
as follows:
Speciesi g = &tpeciesFractioni^g ¦ I PM25 g - BlankMass I
where.
Speciesi g - quarterly peak for a given species "i" (e.g., SO 4)
Species Fracti on^ g = species fraction for species" i"
PM25 g = quarterly peak PM25
BlankMass - assumed monitoring blankmass (e.g., O.f&g / m3)
Note that MATS calculates species fractions from speciated monitors for a limited number
of years. As a result, rather than have species fractions specifically calculated for each
quarter and each year, MATS uses a single set of species fractions to calculate the
weighted quarterly average species concentrations. The species data should be
"representative" of the species fractions that occur during the 5 year FRM monitoring
period selected in MATS.
Abt Associates Inc.
167
June 2010
-------
Daily PM Analysis: Details
Example Calculation Baseline Species Concentrations
MATS multiplies the (non-blank) baseline quarterly peak PM2.5 values. Below is an
example showing the calculation for one year of data. Note, however, that if all five years
and all quarters of baseline quarterly peak PM2.5 (calculated in Step 1) are valid, then
there will be 20 quarters of species values (= 5 years x 4 quarters).
FRM PM2.5 Blank Mass Non-Blank Mass
10.4175 0.5 9.9175
11.4941 0.5 10.9941
13.0624 0.5 12.5624
10.5773 0.5 10.0773
with the species fractions calculated for this particular site:
Quarter
S04
N03
OCMMB
EC
PBW
NH4
Crustal
(Salt)
Q1
0.2382
0.1108
0.3346
0.0201
0.1049
0.1209
0.0705
--
Q2
0.2637
0.0727
0.3178
0.0545
0.1095
0.1273
0.0545
--
Q3
0.1432
0.0557
0.5621
0.0238
0.0959
0.0955
0.0238
--
Q4
0.2580
0.1389
0.2756
0.0396
0.1094
0.1190
0.0595
—
The product is the estimated baseline species concentrations by quarter.
Quarter
S04
N03
OCMMB
EC
PBW
NH4
Crustal
Q1
2.3623
1.0989
3.3184
0.1993
1.0403
1.1990
0.6992
Q2
2.8991
0.7993
3.4939
0.5992
1.2039
1.3996
0.5992
Q3
1.7989
0.6997
7.0613
0.2990
1.2047
1.1997
0.2990
Q4
2.5999
1.3997
2.7773
0.3991
1.1025
1.1992
0.5996
Retained Ammonium Calculation
MATS calculates retained ammonium two different ways. The default approach is to use
interpolated degree of neutralization of sulfate (DON) values from the speciated monitors.
The alternative approach is to use interpolated NH4 values from speciated monitors (e.g.,
STN). In the Species Fractions Calculations Options - Advanced section, you have the
option to choose the approach that you prefer to use. The two approaches are described
here.
Default approach using measured pre-calculated DON, S04, and retained N03
(N03r):
Because of uncertainties in NH4 speciation measurements, by default MATS calculates
ammonium values using the degree of sulfate neutralization (DON). MATS uses pre-
calculated daily DON values that are included in the species data input file ("Species-for-
fractions-xxxx.csv"). The values for DON are calculated from the amount of ammonium
Abt Associates Inc.
168
June 2010
-------
Daily PM Analysis: Details
associated with sulfate (NH4SQ4) as follows:
DON-
dun - ySQ
And the estimated NH4SQ4 is calculated as follows:
NH,^S(A = NH^MeajUKd - 0.29 * ND^reaixd
where 0.29 is the molar ratio of NH4 to N03 and NH4measured and N03retained reflect the
amounts of NH4 and N03 retained on the FRM filter. The amount of NH4SQ4 is not
allowed to exceed the fully neutralized amount of 0.375 multiplied by the estimated sulfate
ion concentration.
MATS then calculates ammonium using interpolated monitor values of DON, S04, and
N03r as follows:
NH, = DON* SO, + 0.29* NOxM
Alternative Approach Using Measured Ammonium. The alternative approach is to use
interpolated NH4 values from STN monitors. This approach has several steps.
First, MATS calculates "adjusted" NH4:
~ {PctEvap *0.29 *{NOlMeasirsd - N03^d)
where the PctEvap factor refers to the percentage of ammonium associated with the
volatilized nitrate that is also lost. (As discussed in the Species Fractions Calculation
Options - Advanced section, this factor is adjustable from 0 to 100 percent.) The default
assumption is that no ammonium is volatilized (0 percent).
Second, MATS calculates NH4 associated with S04:
NH,iSn = -0.29*M)3i„
Third, MATS calculates DON:
DON _NH*,so*/
UUNcaktimed ~ /SO,
Finally, using the same equation as in the default approach, MATS calculates NH4r by
substituting the calculated DON for the interpolated (measured) DON value:
= ^>ONedesjSj^d*SO, + 0.29* hiO^ refcansrf
Abt Associates Inc.
169
June 2010
-------
Daily PM Analysis: Details
Particle Bound Water Calculation
Because ammoniated sulfate and ammonium nitrate are hygroscopic, the retained sulfate
and nitrate mass will include water. Particle bound water (PBW) is estimated using the
Aerosol Inorganic Model (AIM) (Clegg et al, 1998). For computational convenience, a
polynomial regression equation was fit to the calculated water mass from AIM and the
three input values that fed into AIM (sulfate, nitrate and ammonium). AIM was run with
typical FRM filter equilibration conditions of 35% RH and 22 deg C (295 deg K).
MATS calculates particle-bound water (PBW) using levels of S04, N03r, and NH4r as
follows. (Note that this is the same equation that MATS uses to calculate future-year
PBW, the difference being the future-year PBW uses future-year values of S04, N03r, and
NH4r, and here MATS uses base-year values.)
The calculation uses one of two equations, depending on the acidity of the ammoniated
sulfate (represented by DON). S, N, and A in the equations are the relative fraction of
S04, N03r, and NH4r respectively.
S = S04 / (S04 + N03r + NH4r);
N = N03r / (S04 + N03r + NH4r);
A = NH4r / (S04 + N03r + NH4r);
ifDONle 0.225 then
PBW = {595.556
- 1440.584*S
- 1126.488*N
+ 283.907*(S**1.5)
- 13.384*(N**1.5)
- 1486.711*(A**1.5)
+ 764.229*(S**2)
+ 1501.999*(N*S)
+ 451.873*(N**2)
- 185.183*(S**2.5)
- 375.984*(S**1.5)*N
- 16.895*(S**3)
- 65.814*(N**1.5)*S
+ 96.825*(N**2.5)
Abt Associates Inc.
170
June 2010
-------
Daily PM Analysis: Details
+ 83.037*(N**1.5)*(S**1.5)
-4.419*(N**3)
+ 1720.818*(A** 1.5)* S
+ 1220.383*(A**1.5)*N
- 311.496*(A**1.5)*(S**1.5)
+ 148.771*(A**1.5)*(N**1.5)
+ 1151.648*(A**3)} * (S04+N03r+NH4);
ELSE
PBW = {202048.975
-391494.647 *S
-390912.147 *N
+ 442.435 *(S**1.5)
- 155.335 *(N**1.5)
-293406.827 *(A**1.5)
+ 189277.519 *(S**2)
+ 377992.610 *N*S
+ 188636.790 *(N**2)
-447.123 *(S**2.5)
- 507.157 *(S**1.5)*N
- 12.794 *(S**3)
+ 146.221 *(N**1.5)*S
+ 217.197 *(N**2.5)
+ 29.981 *(N**1.5)*(S**1.5)
- 18.649 *(N**3)
+ 216266.951 *(A**1.5)*S
+ 215419.876 *(A**1.5)*N
-621.843 *(A**1.5)*(S**1.5)
+ 239.132 *(A**1.5)*(N**1.5)
Abt Associates Inc.
171
June 2010
-------
Daily PM Analysis: Details
+ 95413.122 *(A**3)} * (S04+N03r+NH4).
Organic Carbon Mass Calculation
Measured organic carbon mass is not directly used in the calculation of species fractions in
MATS because of (1) many uncertainties in estimating carbonaceous mass from carbon
measurements (Turpin & Lim, 2001; Chow et al, 2004) (2) differences in carbon
measurement protocol between urban and rural monitoring locations, (3) a relatively
"bumpy" surface of urban carbon concentrations as derived from urban and rural organic
carbon measurements and (4) lack of carbon measurements at all FRM locations. The
MATS approach estimates carbon by mass balance comparing precisely measured FRM
PM2.5 mass (EPA, 2003) with the sum of its non-carbon components.
Total carbonaceous mass contains both elemental carbon (EC) and organic carbon mass
(OCM). We measure EC from the interpolated STN and IMPROVE monitors, while we
calculate OCM using a mass balance approach — and refer to it as OCMMB. To calculate
OCMMB, we subtract the other estimated retained species (including EC) from the PM2.5
level measured at the FRM site as follows:
OCM^z = PM2S-\ SO< + N03 r^Sid + NH< retsjsed + PBW + Crustal + EC + Blank Mass + {Salt) \
The value for OCMMB could be very small, or even be calculated as negative (if the sum
of the species enclosed in the curly brackets exceeded the FRM PM2.5 monitor value). To
ensure that the OCMMB does not get too small, an OCMMB "mass balance
floor" (default) value is set to 1.0 times the interpolated value of blank-adjusted organic
carbon (OCb). It is also possible that the value of the floor by itself could exceed the FRM
total PM2.5 value. In this case, MATS imposes a (user-adjustable) "ceiling," such that
OCMMB does not exceed a percentage of the total non-blank mass. The default ceiling
value is set to 0.8 or 80% of PM2.5 mass. (You can modify the floor and ceiling
assumptions in the Species Fractions Calculation Options - Advanced window.)
To account for these possibilities, MATS uses the following series of equations to
calculate OCMMB:
0CMmb, Aiifli = NonBiankMass - j SO4 + NOi retajxed + NH^KBixd + PB W + Crustal + EC + Salt \
OCM= Floor* OCb
O C^MB, &2nxediai> = x( ^ MB, Fber'® )
OCMlfSB = Ceiling * NonBlankMass
OCM^
, Final
Min (OCM ^ ceiiHg'O CMm^
Sitermerfiate )
where the "Floor" variable has a default value in MATS of
has a default value in MATS of "0.8".
There are at least two things to note with this approach. (1)
"1.0" and the "Ceiling" variable
When the final OCMMB
Abt Associates Inc.
172
June 2010
-------
Daily PM Analysis: Details
value is equal to the "floor," then the sum of the species will exceed the PM2.5 value at the
FRM monitor. To ensure that the sum of the species just equals the FRM PM2.5 value,
MATS reduces all of the species (except OCMMB) by the same percentage until the sum
of the species just equals the FRM PM2.5 value. (2) When the final OCMMB value is
equal to the "ceiling," then the sum of the species will be less than the PM2.5 value at the
FRM monitor. In that case, MATS increases all of the species by the same percentage
until the sum of the species just equals the FRM PM2.5 value
Blank Mass Assumption
The field blank typically has a value of between 0.3 and 0.5 ug/m3, which appears to result
from contamination of the FRM filter. For calculating retained PM2.5, MATS uses a
default blank mass value of 0.5 ug/m3. If desired, you can change the default blank mass
value at the Species Fractions Calculations Options - Advanced window.
7.1.2.3 Step 3: Forecasted Quarterly Peak Species Calculation
To calculate forecasted quarterly peak species values, MATS uses each valid baseline
quarterly peak species values and both baseline and forecasted (e.g., 2020) air quality
modeling. If five years of species estimates are available and if all quarters are valid, then
this calculation gives 20 quarterly estimates for each species. Because this calculation
involves modeling data from two different years, this is referred to as a "temporal
adjustment."
The forecasted weighted quarterly peak for each species is calculated by multiplying the
baseline quarterly peak values for each monitor species with the ratio of the modeling data.
This process gives quarterly peak values of six species: S04, N03, OCM, EC, crustal
material, and NH4. The form of the equation is as follows:
r, . n M°dsh,jlitiLreQ
l^PsclssiJmreY, Q ~ ^Peclesi,iax Y, Q ' , ,
1V10asiiJ}iLxQ
where
Speciesi g ~ Estimated forecasted species" in peakinyear Y and quarter Q
3peciesijj)axY g = Monitored baseline species "i" peakinyear Y and quarter Q
Modelj^^g = Modeled forecasted species " z" peak in quarter Q
Modeling = Modeledbaseline species "i" peakinquarter Q
In other words, baseline species concentrations are assumed to change in the same
proportion as the model data in the same location from the baseline to the forecasted.
These proportions, called relative response factors (RRF), are simply the ratio of the
modeled forecasted species to the modeled baseline species.
Abt Associates Inc.
173
June 2010
-------
Daily PM Analysis: Details
McdsL
— ujatuwy
1,8 Model^g
where
RRFg = reia five re span se factor for quarter Q
MATS calculates RRFs for each quarter for each of six species: S04, N03, OCM, EC,
crustal material, and (optionally) NH4. (Additional information on the calculation of the
RRFs can be found in the Model Data Options section.) The calculation of forecasted
quarterly peaks species concentrations can be rewritten as:
SpeneSiJumreB = 3psciesi^eB ¦ RRFi,e
Additional calculations are needed to estimate future-year quarterly peak values of NH4
and particle-bound water (PBW), which is calculated using forecasted levels of NH4, N03,
and S04. (Details on the PBW calculation can be found here.')
Recall that the default base year NH4 calculation is as follows:
m, = DON* SO, + 0.29*
MATS can calculate the future year NH4 concentration using modeled NH4 RRFs, or by
using the the base year DON value combined with future year S04 and N03 values
(default approach) as follows:
¦^^4, future = DON* SO ^future + Q.29*NO^
The option for choosing which approach to use for calculating future NH4 is given in the
PM2.5 Calculation Options window. Finally, note that PBW is calculated after future year
NH4, using the previously identified water equation and future year concentrations of
NH4, S04, and N03.
Example Calculation Forecasted Species Concentrations Using RRFs
MATS multiplies the baseline species concentrations for S04, N03, OCMMB, EC, and
Crustal. Below is an example showing the calculation for one year of data. Note,
however, that if all five years and all quarters of baseline quarterly peak PM2.5 (calculated
in Step 1) are valid, then there will be 20 quarters of species values (= 5 years x 4
quarters).
Quarter
S04
N03
OCMMB
EC
PBW
NH4
Crustal
Q1
2.3623
1.0989
3.3184
0.1993
1.0403
1.1990
0.6992
Q2
2.8991
0.7993
3.4939
0.5992
1.2039
1.3996
0.5992
Q3
1.7989
0.6997
7.0613
0.2990
1.2047
1.1997
0.2990
Q4
2.5999
1.3997
2.7773
0.3991
1.1025
1.1992
0.5996
Abt Associates Inc.
174
June 2010
-------
Daily PM Analysis: Details
with the RRFs for each quarter and species:
Quarter
S04
N03
OCMMB
EC
PBW
NH4
Crustal
Q1
0.9737
0.9873
0.9636
0.9872
0.9808
Q2
0.9898
0.9991
0.9979
0.9917
0.9891
Q3
0.9784
0.9775
0.9875
0.9944
0.9759
Q4
0.9700
0.9800
0.9761
0.9843
0.9818
The product is the forecasted species concentrations by quarter.
Quarter
S04
N03
OCMMB
EC
PBW
NH4
Crustal
Q1
2.3003
1.0849
3.1975
0.1968
--
--
0.6858
Q2
2.8695
0.7985
3.4865
0.5942
--
--
0.5927
Q3
1.7600
0.6840
6.9728
0.2973
--
--
0.2918
Q4
2.5219
1.3717
2.7109
0.3928
—
—
0.5887
As discussed earlier, the calculation for PBW and the (default) calculation for NH4 does
not directly involve using RRFs.
7.1.2.4 Step 4: Forecasted Design Value Calculation
The forecasted quarterly peaks for the species are then added together to get a estimated
forecasted quarterly peak PM2.5 values for each year of monitor data:
PM2$ y q - SO4 y q+ NQ y q + OCMy q + ECy q + CrustalY g + NHA Y q + PBWY g
These forecasted quarterly PM2.5 concentration represent a potential daily design value.
This procedure is repeated for each of the years of monitoring data (e.g., 2000-2004). The
highest forecasted quarterly peak value for each year at each monitor is considered to be
the estimated 98th percentile value for that year.
PM25 v = Max i PM25 v g i
The estimated 98th percentile values for each of the 5 years are averaged over 3 year
intervals (e.g., 2000-2002, 2001-2003, 2002-2004):
PM.
25, jenlt £72002
^^2.5, 2COO + PM2 J,2001 + ^'^2 J, 2®2
PM,
2 5, peak DV20(B
PM-
2j,^atDV2004
PM2i2mi + PM2X
2032
+ PM2
.5,2003
P^2.5,2002 + PM2 5,2003 + *-P^2J,20 04
Abt Associates Inc.
175
June 2010
-------
Daily PM Analysis: Details
The design value is then the average of the three peak design values, creating 5-year
weighted average for each monitor:
n , PM2 m 3]02 + PM2 ^ DY + PM2 ^ Dy 3004
P"^25,feai.DV ~ 3
where
PM25 peayzv ~ peak PM2 c design value.
The output file containing the results of this Standard Analysis are described here.
Example Calculation Forecasted Design Values
Using hypothetical data from the previous example calculation plus PBW values and
assumed blank mass, MATS sums the species to get quarterly peak PM2.5 values for each
year.
Forecasted Species
Concentrations
Year
Quarter
S04
N03
OCMMB
EC
PBW
NH4
Crustal
Blank
Mass
PM2.5
2000
Q1
2.353
1.141
3.358
0.207
0.500
1.214
0.721
0.500
9.99
Q2
2.638
1.274
3.837
0.230
0.400
1.349
0.802
0.500
11.03
Q3
2.964
1.416
4.315
0.262
0.600
1.533
0.900
0.500
12.49
Q4
2.379
1.150
3.454
0.210
0.700
1.253
0.733
0.500
10.38
2001
Q1
3.553
1.141
3.103
0.083
0.515
1.922
0.768
0.500
11.59
Q2
1.905
1.780
4.848
0.136
0.397
1.505
0.589
0.500
11.66
Q3
3.509
1.182
1.621
0.414
0.561
1.635
0.256
0.500
9.68
Q4
1.725
0.312
2.910
0.361
1.112
2.090
0.751
0.500
9.76
2002
Q1
2.958
1.660
4.631
0.297
0.313
1.582
0.991
0.500
12.93
Q2
1.966
1.902
3.547
0.204
0.253
2.213
0.863
0.500
11.45
Q3
2.724
2.246
5.628
0.184
0.277
1.365
1.492
0.500
14.42
Q4
0.434
0.465
4.560
0.253
0.643
0.987
1.006
0.500
8.85
2003
Q1
2.291
1.243
5.307
0.223
0.711
1.665
0.250
0.500
12.19
Q2
2.003
0.731
0.950
0.289
0.428
1.263
0.848
0.500
7.01
Q3
4.013
0.816
5.264
0.414
0.605
2.014
1.247
0.500
14.87
Q4
2.311
0.667
3.726
0.079
0.622
0.443
1.008
0.500
9.35
2004
Q1
4.129
0.255
3.360
0.200
0.949
2.253
0.627
0.500
12.27
Q2
1.854
1.449
2.867
0.228
0.090
1.160
1.153
0.500
9.30
Q3
3.275
1.058
4.352
0.201
0.679
1.252
1.230
0.500
12.55
Q4
3.237
0.503
5.251
0.291
0.770
1.910
0.878
0.500
13.34
Then MATS selects the highest quarterly peak in each year (e.g., 12.49 for year 2000 in the
example here):
Year Annual Peak
2000 12.49
2001 11.66
2002 14.42
2003 14.87
2004 13.34
These annual peaks are then averaged for each three-year period, and then MATS averages
Abt Associates Inc.
176
June 2010
-------
Daily PM Analysis: Details
these three-year averages to get the forecasted peak design value:
Year Design Values
2002 12.86
2003 13.65
2004 14.21
Forecasted 13.6
Design Value
7.1.2.5 Output Description
The output file is named "Daily PM25 Point.csv" with the Scenario Name appended at the
beginning (e.g., "Example Daily PM Daily PM25 Point.csv"). The table below describes
the variables in the output file for the Standard Analysis.
Note the following:
The RRF variables in this file are not the actual RRFs used to calculate future peak
PM2.5 and PM2.5 species. They are the resultant annual RRFs calculated by dividing
the future peak concentrations (in this file) by the base year peak concentrations (in this
file). The actual RRFs are calculated on a quarterly basis and are contained in the
quarterly peak output files. There are no quarterly RRFs for water or NH4 (if DON,
N03, and S04 are used to calculate NH4).
The baseline design value ("b_pm25_d_DV") and future design value ("f
_pm25_d_DV") are calculated independently, so the peak the quarters used for each
may differ. Also, note that these values are truncated with one digit after the decimal
point.
Output file name: "Scenario Name + Daily PM25 Point"
Variable Description
Jd The ID is a unique name for each monitor in a particular location. The default
value is the AIRS ID. (This is a character variable.)
_type FRM data
_STATE_NAME State name. (This is a character variable.)
_COUNTY_NAM County name. (This is a character variable.)
E
monitorjat
monitorjong
monitor_gridcell
b_pm25_d_DV
f_pm25_d_DV
b blank mass
Latitude at the monitor site in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are negative.
Longitude at the monitor site in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g., United
States) are negative.
Identifier of grid cell closest to the monitor
Base year 5 year weighted average PM2.5 24-hr average (daily) design value
Future year 5 year weighted average PM2.5 24-hr average (daily) design value
Base year blank mass concentration (ug/m3)
Abt Associates Inc.
177
June 2010
-------
Daily PM Analysis: Details
b_crustal_mass
b_EC_mass
b_NH4_mass
b_Ocmb_mass
b_S04_mass
b_N03_mass
b_water_mass
b_salt_mass
f_blank_mass
f_crustal_mass
f_EC_mass
f_NH4_mass
f_Ocmb_mass
f_S04_mass
f_N03_mass
f_water_mass
f_salt_mass
rrf_crustal
rrf_ec
rrf_nh4
rrf_oc
rrf_so4
rrf_no3
rrf_water_mass
rrf salt
Base year crustal mass concentration (ug/m3)
Base year elemental carbon mass concentration (ug/m3)
Base year ammonium mass concentration (ug/m3)
Base year organic carbon mass (by difference) concentration (ug/m3)
Base year sulfate ion mass concentration (ug/m3)
Base year nitrate ion mass concentration (ug/m3)
Base year water mass concentration (ug/m3)
Base year salt mass concentration (ug/m3)
Future year blank mass concentration (ug/m3)
Future year crustal mass concentration (ug/m3)
Future year elemental carbon mass concentration (ug/m3)
Future year ammonium mass concentration (ug/m3)
Future year organic carbon mass (by difference) concentration (ug/m3)
Future year sulfate ion mass concentration (ug/m3)
Future year nitrate ion mass concentration (ug/m3)
Future year water mass concentration (ug/m3)
Future year salt mass concentration (ug/m3)
Resultant annual relative response factor- Crustal Mass
Resultant annual response factor- Elemental Carbon Mass
Resultant annual relative response factor- Ammonium Mass.
Resultant annual relative response factor- Organic Carbon Mass
Resultant annual relative response factor- Sulfate Mass
Resultant annual relative response factor- Nitrate Mass
Resultant annual relative response factor- Water Mass
Resultant annual relative response factor- Salt Mass (set equal to 1 if modeled
salt is not used)
7.1.3 Quarterly Peak Model Data
The Quarterly Peak Model Data option gives you the option of creating a small, reusable
file with quarterly peak values from a much larger file with daily values. Since daily
PM2.5 MATS works with quarterly peak values there is no loss of precision. To save
time, it is possible to use daily values only for an initial run with MATS and check this
Quarterly Peak Model Data option. Then for subsequent runs (that use the same
modeled data), use the quarterly file that MATS generates. However, this will only work
for subsequent MATS runs that use exactly the same base and future year photochemical
model data (such as sensitivity runs that test the various ambient data settings in MATS).
Alternatively, you can generate a baseline and future quarterly average model file outside
of MATS (using a program such as SAS, STAT A, etc), and then load these quarterly files
Abt Associates Inc.
178
June 2010
-------
Daily PM Analysis: Details
into MATS, bypassing the use of any daily model data in MATS. The format of the
quarterly file is described below.
MATS generates baseline and future model files for the complete model domain:
Base-year file: "Baseline Quarterly Peak Model Data.csv" with the Scenario Name (e.g.,
Example Daily PM) appended at the beginning (e.g., "Example Daily PM Baseline
Quarterly Peak Model Data.csv").
Future year file: "Future Quarterly Peak Model Data.csv."
MATS also produces a set of "used" quarterly peak model files (base and future). These
have the same format as the complete quarterly peak model data files, but only contain data
for the model grid cells that were used in the MATS point calculations. These files can
also be re-used and are also useful for QA purposes.
Base-year file: " Used Baseline Quarterly Peak Model Data Point.csv."
Future year file: "Used Future Quarterly Peak Model Data Point.csv."
The format of all four of these model files is the same. The table below describes the
variables in the Quarterly Peak Model Data file.
Output file name: "Scenario Name + (Used) Baseline/Future Quarterly Peak Model
Data (Point)"
Variable Description
Jd The ID is a unique identifier for each model grid cell. The default value is the
column identifier multiplied by 1000 plus the row. (This is a character variable.)
Jype
lat Latitude at the grid cell centroid in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are negative.
long Longitude at the grid cell centroid in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g., United
States) are negative,
date Year and Quarter (01= 1st quarter, 04= 2nd quarter, etc.)
crustal Crustal PM
nh4 Ammonium PM
so4 Sulfate PM
ec Elemental Carbon
no3 Nitrate PM
oc Organic carbon PM
pm25 PM2.5 mass (only used to gradient adjust PM2.5 for gradient adjusted spatial
fields)
cm Coarse PM (ug/m3) (only used for visbility calculations)
Abt Associates Inc.
179
June 2010
-------
Daily PM Analysis: Details
7.1.4 Species Fractions
Species fractions are simply the fraction of quarterly average PM2.5 at a given monitor
attributable to seven (and potentially eight) species: nitrate (N03), sulfate (S04), organic
carbon (OC), crustal, elemental carbon (EC), ammonium (NH4), and particle-bound water
(PBW). (And pending data availability, an eighth species, salt, can be included as well, the
default MATS species files include salt data. But salt is an optional species on the model
files. If base and future year modeled salt data is supplied, a salt RRF will be calculated.
If there is no salt data in the model files, then the salt RRF will be set to 1.)
When you check the Species Fractions option, you get a reusable file with species
fractions for each FRM monitor. Making the file reusable allows you to generate
consistent results and, perhaps most importantly, allows the same file to be used by
different MATS users.
7.1.4.1 Species Fractions Calculation
After calculating the ambient level of S04, N03, OCMMB, EC, PBW, NH4, and crustal
(as described in Step 2 of the Standard Analysis). MATS then divides these ambient levels
by the non-blank mass of PM2.5. To get non-blank PM2.5, MATS subtracts the blank
mass from the FRM PM2.5 value. MATS then divides each of the species (except blank
mass) by non-blank PM2.5.
Example Calculation of Species Fractions
The table below gives an example of theses calculations. The fraction is calculated by
dividing mass (ug/m3) by the Non-blank Mass. Note that salt is optional.
Units FRM Blank Non-BIa S04 N03 OCMM EC PBW NH4 Crusta (Salt)
PM2.5 Mass nk Mass B I
Concentrati 10.9175 0.5 10.4175 2.41601.1555 3.4847 0.2101 1.15551.26050.7353 -
on
Fraction 0.23190.1109 0.3345 0.0202 0.1109 0.1210 0.0706 -
7.1.4.2 Output Description
The output file is named "Quarterly Peak Spec Frac Point.csv" with the Scenario Name
appended at the beginning (e.g., "Example Daily PM Quarterly Peak Spec Frac Point.csv
"). The table below describes the variables in the file.
The interpolated variables (starting with i xxx) are created when MATS is run, but are not
needed when re-using a fractions file. They are also not needed when running with a user
generated fractions file.
Output file name: "Scenario Name + Quarterly Peak Spec Frac Point"
Variable Description
Jd The ID is a unique name for each monitor in a particular location. The
default value is the AIRS ID. (This is a character variable.)
Abt Associates Inc.
180
June 2010
-------
Daily PM Analysis: Details
_STATE_NAME
_COUNTY_NAME
monitorjat
monitorjong
quarter
pm25_mass_frac
fcr
fee
fnh4
focm
fso4
fno3
fwater
fsalt
blank_mass
don
_so4
_no3r
_ocb
_ec
_crustal
_don
_nh4
_no3
i salt
State name. (This is a character variable.)
County name. (This is a character variable.)
Latitude at the monitor site in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are
negative.
Longitude at the monitor site in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g.,
United States) are negative.
Quarter
PM2.5 mass used to calculate species fractions
Crustal fraction of PM2.5 mass
Elemental carbon fraction of PM2.5 mass
Ammonium fraction of PM2.5 mass
Organic carbon fraction of PM2.5 mass
Sulfate ion fraction of PM2.5 mass
Nitrate ion fraction of PM2.5 mass
Water fraction of PM2.5 mass
Salt fraction of PM2.5 mass
Blank mass
Degree of neutralization of sulfate used to calculate ammonium mass
(0.000 - 0.375)
Interpolated sulfate ion
Interpolated retained nitrate ion
Interpolated blank-adjusted organic carbon
Interpolated elemental carbon
Interpolated crustal
Interpolated degree of neutralization of sulfate (DON).
Interpolated ammonium
Interpolated nitrate ion (only used to calculate volatilized ammonium; if
option is selected)
Interpolated salt
Note:
i ocb is only used to to calculate the OCMmb "floor".
i_nh4 is not used if DON is used to calculate the ammonium concentration (and
fraction).
i_no3 is only used to calculate the "volatilized ammonium", if the option is selected (not
used by default).
Abt Associates Inc.
181
June 2010
-------
Daily PM Analysis: Details
7.2 Output Choice - Advanced
In the Output Choice Advanced window, MATS lets you choose various files grouped
under the heading Miscellaneous Output. These files are generally used for QA and are
described below.
7.2.1 Miscellaneous Output
MATS gives you ability to generate a number of specialized files for quality assurance,
identify the monitor in each county with the highest forecast, and other reasons. These
files include:
Quarterly average files. The MATS default is to output the peak design values, which
are based on a weighted average of five years of data. However, checking this box will
output more detailed quarterly calculations (the quarterly peak calculations are the basis
of all of the peak design value calculations). There are two quarterly average files. The
first contains quarterly peak data for years and quarters, and contains baseline PM2.5
and species concentrations, RRFs, forecasted PM2.5 and species concentrations. The
"true" RRFs and PM2.5 and species concentrations are found in this file. A second file
has the peak quarter and PM2.5 value in the baseline and future for each year at each
monitor. Both of these quarterly files are important because all of the basic PM2.5
calculations in MATS occur on a quarterly basis.
High county sites. The MATS default is to output the point results for all FRM sites.
Checking this box will also create a file which contains only the single highest monitor
Abt Associates Inc.
182
June 2010
-------
Daily PM Analysis: Details
in each county (based on the highest future-year value). This dataset is a subset of the
all sites file.
Quarterly average speciated monitors. This file contains the raw quarterly peak
speciated data at STN and IMPROVE sites that MATS uses to do interpolations (to
calculate species fractions). This is the subset of species data that MATS uses for each
particular scenario (based on the MATS inputs and configuration settings).
Neighbor files. The neighbor files contain the "nearest neighbor" data for the VNA
interpolation scheme. The data includes the distance to neighbor monitors and weights
used to do the interpolations. There is information for each FRM monitor for each
quarter and for each species.
7.2.1.1 Quarterly Average Files
The Quarterly Average Files provide intermediate calculations performed by MATS.
There are two files generated by checking this option:
"Daily All Years All Quarters PM25 Point". This file has baseline and future peak
values for PM2.5 and its constituent species. In addition it gives the speciated RRFs.
"Daily All Years High Quarters PM25 Point". This file identifies the peak quarter and
PM2.5 value in the baseline and future for each year at each monitor. Note that the
baseline and future quarters may differ.
When generating these files, MATS appends the Scenario Name at the beginning (e.g., "
Example Daily PM Daily All Years All Quarters PM25 Point, csv"). The variables in each
file are described below.
Output file name:
Variable
Jd
Jype
_STATE_NAME
_COUNTY_NAME
monitorjat
monitorjong
monitor_gridcell
year
quarter
"Scenario Name + Daily All Years All Quarters PM25 Point"
Description
The ID is a unique name for each monitor in a particular location. The
default value is the AIRS ID. (This is a character variable.)
FRM data
State name. (This is a character variable.)
County name. (This is a character variable.)
Latitude at the monitor site in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are
negative.
Longitude at the monitor site in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g.,
United States) are negative.
Identifier of grid cell closest to the monitor
Year (up to 5 years)
Quarter
Abt Associates Inc.
183
June 2010
-------
Daily PM Analysis: Details
b_pm25_d_q_conc
f_pm25_d_q_conc
b_blank_mass_q
b_crustal_mass_q
b_EC_mass_q
b_NH4_mass_q
b_Ocmb_mass_q
b_S04_mass_q
b_N03_mass_q
b_water_mass_q
b_salt_mass_q
f_blank_mass_q
f_crustal_mass_q
f_EC_mass_q
f_NH4_mass_q
f_Ocmb_mass_q
f_S04_mass_q
f_N03_mass_q
f_water_mass_q
f_salt_mass_q
rrf_crustal_q
rrf_ec_q
rrf_nh4_q
rrf_oc_q
rrf_so4_q
rrf_no3_q
rrf_salt_q
Base year peak PM2.5 concentration (high concentration for each
quarter)
Future year peak PM2.5 concentration (high concentration for each
quarter)
Base year blank mass concentration (ug/m3)
Base year crustal mass concentration (ug/m3)
Base year elemental carbon mass concentration (ug/m3)
Base year ammonium mass concentration (ug/m3)
Base year organic carbon mass (by difference) concentration (ug/m3)
Base year sulfate ion mass concentration (ug/m3)
Base year nitrate ion mass concentration (ug/m3)
Base year water mass concentration (ug/m3)
Base year salt mass concentration (ug/m3)
Future year blank mass concentration (ug/m3)
Future year crustal mass concentration (ug/m3)
Future year elemental carbon mass concentration (ug/m3)
Future year ammonium mass concentration (ug/m3)
Future year organic carbon mass (by difference) concentration (ug/m3)
Future year sulfate ion mass concentration (ug/m3)
Future year nitrate ion mass concentration (ug/m3)
Future year water mass concentration (ug/m3)
Future year salt mass concentration (ug/m3)
Relative response factor- Crusatal Mass
Relative response factor- Elemental Carbon Mass
Relative response factor- Ammonium Mass. (Not used if future year
ammonium is calculated using base year DON values)
Relative response factor- Organic Carbon Mass
Relative response factor- Sulfate Mass
Relative response factor- Nitrate Mass
Relative response factor- Salt Mass
Output file name: "Scenario Name + Daily All Years High Quarters PM25 Point"
Variable Description
Jd The ID is a unique name for each monitor in a particular location. The
default value is the AIRS ID. (This is a character variable.)
Jype
_STATE_NAME
COUNTY NAME
FRM data
State name. (This is a character variable.)
County name. (This is a character variable.)
Abt Associates Inc.
184
June 2010
-------
Daily PM Analysis: Details
monitorjat
monitorjong
monitor_gridcell
year
b_high_quarter
b_pm25_d_q_conc
f_high_quarter
f_pm25_d_q_conc
Latitude at the monitor site in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are
negative.
Longitude at the monitor site in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g.,
United States) are negative.
Identifier of grid cell closest to the monitor
Year (up to 5 years)
Base high quarter number
Base year peak PM2.5 concentration (high quarter for each year)
Future high quarter number
Future year peak PM2.5 concentration (high quarter for each year)
7.2.1.2 Output Description - High County Sites
In this file, MATS reports the monitor with the highest forecasted peak PM2.5 design
value in each county. The name of this file is "High County Sites" with the Scenario
Name appended at the beginning (e.g., "Example Daily PM — High County Sites.csv").
The table below describes the variables in the output file.
Output file name: "Scenario Name + High County Sites"
Variable Description
Jd The ID is a unique name for each monitor in a particular location. The default
value is the AIRS ID. (This is a character variable.)
_type FRM data
_STATE_NAME State name. (This is a character variable.)
_COUNTY_NAME County name. (This is a character variable.)
monitorjat
monitorjong
monitor_gridcell
b_pm25_d_DV
f_pm25_d_DV
b_blank_mass
b_crustal_mass
b_EC_mass
b_NH4_mass
b_Ocmb_mass
b SQ4 mass
Latitude at the monitor site in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are negative.
Longitude at the monitor site in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g., United
States) are negative.
Identifier of grid cell closest to the monitor
Base year 5 year weighted average PM2.5 24-hr average (daily) design value
Future year 5 year weighted average PM2.5 24-hr average (daily) design value
Base year blank mass concentration (ug/m3)
Base year crustal mass concentration (ug/m3)
Base year elemental carbon mass concentration (ug/m3)
Base year ammonium mass concentration (ug/m3)
Base year organic carbon mass (by difference) concentration (ug/m3)
Base year sulfate ion mass concentration (ug/m3)
Abt Associates Inc.
185
June 2010
-------
Daily PM Analysis: Details
b_N03_mass
b_water_mass
b_salt_mass
f_blank_mass
f_crustal_mass
f_EC_mass
f_NH4_mass
f_Ocmb_mass
f_S04_mass
f_N03_mass
f_water_mass
f_salt_mass
rrf_crustal
rrf_ec
rrf_nh4
rrf_oc
rrf_so4
rrf_no3
rrf_water_mass
rrf salt
Base year nitrate ion mass concentration (ug/m3)
Base year water mass concentration (ug/m3)
Base year salt mass concentration (ug/m3)
Future year blank mass concentration (ug/m3)
Future year crustal mass concentration (ug/m3)
Future year elemental carbon mass concentration (ug/m3)
Future year ammonium mass concentration (ug/m3)
Future year organic carbon mass (by difference) concentration (ug/m3)
Future year sulfate ion mass concentration (ug/m3)
Future year nitrate ion mass concentration (ug/m3)
Future year water mass concentration (ug/m3)
Future year salt mass concentration (ug/m3)
Resultant annual relative response factor- Crustal Mass
Resultant annual response factor- Elemental Carbon Mass
Resultant annual relative response factor- Ammonium Mass.
Resultant annual relative response factor- Organic Carbon Mass
Resultant annual relative response factor- Sulfate Mass
Resultant annual relative response factor- Nitrate Mass
Resultant annual relative response factor- Water Mass
Resultant annual relative response factor- Salt Mass (set equal to 1 if modeled
salt is not used)
7.2.1.3 Output Description - Quarterly Average Speciated Monitors
In this file, MATS reports the quarterly peak values at the speciated monitors. The
speciated data is split into two files:
"Quarterly Peak Speciated Monitors". This has crustal, EC, OC, S04, N03, retained
N03, and salt quarterly peak speciated values.
"Quarterly Peak NH4_DON Monitors". This has NH4 and DON quarterly peak
speciated values.
MATS generates these files with the Scenario Name appended at the beginning (e.g., "
Example Daily PM Quarterly Peak Speciated Monitors, csv"). The tables below describe
the variables in each file.
Output file name: "Scenario Name + Quarterly Peak Speciated Monitors"
Variable Description
Jd IMPROVE/STN/SPECTRE Site Code
_type Monitor type (e.g., STN)
Abt Associates Inc.
186
June 2010
-------
Daily PM Analysis: Details
monitorjat
monitorjong
monitor_gridcell
quarter
b_crustal_mass
b_EC_mass
b_OCB_mass
b_S04_mass
b_N03_mass
b_no3r_mass
b salt mass
Latitude at the monitor site in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are
negative.
Longitude at the monitor site in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g.,
United States) are negative.
Identifier of grid cell closest to the monitor
Quarter
Base year crustal mass concentration (ug/m3)
Base year elemental carbon mass concentration (ug/m3)
Base year organic carbon mass blank adjusted (ug/m3)
Base year sulfate ion mass concentration (ug/m3)
Base year nitrate ion mass concentration (ug/m3)
Base year retained nitrate ion mass concentration (ug/m3)
Base year salt mass concentration (ug/m3)
Output file name:
Variable
Jd
Jype
monitorjat
monitorjong
quarter
b_NH4_mass
b_DON
monitor_gridcell
'Scenario Name + Quarterly Peak NH4 & DON Monitors"
Description
IMPROVE/STN/SPECTRE Site Code
Monitor type (e.g., STN)
Latitude at the monitor site in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are
negative.
Longitude at the monitor site in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g.,
United States) are negative.
Quarter
Base year ammonium mass concentration (ug/m3)
Base year degree of neutralization (DON)
Identifier of grid cell closest to the monitor
7.2.1.4 Neighbor Files
MATS calculates the nearby monitors or "neighbors" separately for each species when
interpolating to FRM monitors ("points"). In this file, MATS reports the neighbors
involved in interpolating species values to the FRM monitor sites.
The name of this file is "Neighbor File Point" with the Scenario Name appended at the
beginning (e.g., "Example Daily PM Neighbor File Point.csv"). The table below describes
the variables in the output file.
Output file name: "Scenario Name + Neighbor File Point"
Variable Description
Abt Associates Inc.
187
June 2010
-------
Daily PM Analysis: Details
Jd
_state_name
_county_name
monitorjat
monitorjong
monitor_gridcell
quarter
_neighbor
neighbor_gridcell
distance
weightdistance
weightdistancesquared
pollutant
The ID is a unique name for each monitor in a particular location. The
default value is the AIRS ID. (This is a character variable.)
State name. (This is a character variable.)
County name. (This is a character variable.)
Latitude at the monitor site in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are
negative.
Longitude at the monitor site in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g.,
United States) are negative.
Identifier of grid cell closest to the monitor
Quarter
IMPROVE/STN Site Code. (This is a character variable.)
Identifier of grid cell closest to the neighbor
Distance in kilometers from FRM monitor site and
IMPROVE/STN/SPECTRE neighbor.
Inverse-distance weight
Inverse-distance-squared weight
Pollutant (e.g., S04). Note interpolation approach can vary by pollutant
7.3 Data Input
In the Data Input window, you specify the MATS input files that are used in each scenario.
There are three main types of files which must be specified. These include ambient
PM2.5 species data (STN and IMPROVE), ambient total PM2.5 data (FRM and
IMPROVE), and gridded model output data (e.g. CMAQ or CAMx data).
There is specific terminology that is used on the Data Input page. "Official" data refers to
PM2.5 FRM data that can be used to determine official design values for compliance
purposes (comparison to the NAAQS). Other datasets which may not have rigid regulatory
significance are sometimes referred to as "unofficial" data. The individual input file
choices are explained below.
Abt Associates Inc.
188
June 2010
-------
Daily PM Analysis: Details
Daily PM Analysis
¦ Choose Desired Output
¦ Output Choices-Advanced
Data Input
¦ Species Fractions Options
¦ Species Fractions - Advanced
¦ PM 2.5 Calculation 0 ptions
¦ Model Data Options
¦ Final Output and Check
Data Input
Species Data
® ISoecies Monitor Data File! |C:\Prograrn FilesVAbt Associates\MATS\SampleData\Species-f |
O Species Fractions File [
| spatial field
PM2.5 Monitor Data
Unofficial Dailv Averaae PM2.5 Data File ffor All Species Fractions & PM2.5 Spatial Fielcfi
| C: \Program Files\Abt Associates\MAT S \S ampleD ata\PM 25-for-fractions-0206-v2. csv
Official Daily Average FRM Data File (for PM2.5 Point Calculations)
IC: \Program FilesSAbt AssociatesSMAT S \S ampleD ata\24-hr-official-FR M -99-07-v2. csv
Model Data
® Daily model data input O Quarterly peak model data input
B aseline File | C: \Program FilesSAbt Associates\MAT S \S ampleD ata\2002cc_E U S_PM 25_sub. csv
Forecast File | C: \Program FilesSAbt Associates\MAT S \S ampleD ata\2020cc_E U S_PM 25_sub. csv —|
Location Helper |
< Back Next > Cancel
Species Data. MATS needs ambient PM2.5 species data to calculate species
concentrations at FRM monitoring sites. Users have a choice of supplying a "Species
Monitor Data File" or a "Species Fractions File".
Species Monitor Data File. The default is to provide a species monitor data file.
MATS is populated with daily average species data from STN and IMPROVE sites
across the country. However, users can also provide their own ambient data file.
MATS uses the daily average species data to calculate species fractions at each FRM
monitor. The species fraction data is combined with the "unofficial daily average
PM2.5 data" to calculate species concentrations. The default MATS species data file
contains all available data. However, there is a data flag to indicate site days that are
recommended to be removed from the species fractions calculations. This is not
necessarily the same data flags that have been identified by State agencies. MATS has
incorporated flagging routines that remove data that are considered to be outliers
and/or incomplete data. (A description of the flags is provided in the section on
Species Fractions Calculation Options.)
Species Fraction File. Alternatively, the user can choose to use a pre-calculated
species fractions file which contains quarterly species information for the FRM
monitors of interest. MATS can also re-use a species fractions file that has previously
been generated by MATS. To re-use a previously created fractions file, simply supply
the correct path to the file. (The calculation of species fractions is discussed here.)
PM2.5 Monitor Data. MATS uses both "official" and "unofficial" data in its
Abt Associates Inc.
189
June 2010
-------
Daily PM Analysis: Details
calculations.
Unofficial Daily Average PM2.5 Data File. The "unofficial daily average PM2.5" file
contains the PM2.5 data that is needed to calculate species fractions. It is used in
combination with the "species monitor data file" from above. The unofficial daily
average PM2.5 file is not needed if the user supplies a pre-calculated species fractions
file.
Similar to the species monitor data file from above, the "unofficial" PM2.5 data file
contains a data flag to indicate site days that are recommended to be removed from the
species fractions calculations. The flagged data is matched between the species file and
the PM2.5 file so that the same site days are removed. However, the PM2.5 data file
contains additional data (sites that don't contain speciation measurements) and therefore
has additional flagged site days. These are not the same data flags that have been
identified by State agencies. MATS has incorporated flagging routines that remove data
that are considered to be outliers and/or incomplete data. (A description of the flags is
provided in the section on Species Fractions Calculation Options.^ The user is free to
unflag existing data or add flags as necessary and appropriate.
Official Daily Average FRM Data File. The "official daily average file" contains all
of the "official" daily FRM data that has been used to calculate daily PM2.5 design
values. It is used to calculate design values and 5 year weighted-average design values
as part of the attainment test.
The default data file in MATS was created by the Air Quality Analysis Group within
OAQPS. In most cases, the data should not be altered, however in some cases (e.g.
sensitivity analyses) there may be a need to add or remove data.
Model Data The "model data" refers to gridded model output from models such as
CMAQ or CAMx. The user can choose either daily model data input or quarterly model
data input (which would be quarterly peak values calculated from the daily model data).
Either will work. The default setting is daily average data. Recall that MATS can
generate quarterly peak model data (which can then be re-used in subsequent MATS
runs).
Model data must be selected for all MATS runs. The size of the modeling grid defines
the outputs for point estimates. For point estimates, MATS will output the results for all
specified monitors within the domain.
Note that you need to specify both a Baseline File and a Forecast File. The baseline file
should be consistent with the historical monitor data that you use, and the forecast year
is the future-year of interest.
7.3.1 Species Data Input
The species data may be in form of monitor data (specified below) that MATS then uses to
Abt Associates Inc.
190
June 2010
-------
Daily PM Analysis: Details
calculate species fractions, or it may be in the form of species fractions directly (specified
here).
Monitor data should be in the form of a simple text file. The first row specifies the
frequency of the data (e.g., day). The second row presents comma-separated variable
names. The third row begins the data values. Below is an example of the monitor data file
format and descriptions of the variables in the file.
Format of Speciated Monitor Data
Species-for-fractions-0206-v2.csv - WordPad
File Edit View Insert Format Help
BBS
~ & Bk 04
%
Courier Hew
10 v
Western
B Z U
Day
_id,lat,long,
"010050002",3
"010050002",3
"010050002",3
"010050002",3
"010050002",3
"010050002",3
"010050002",3
"010050002",3
"010050002",3
"010050002",3
"010050002",3
"010050002",3
"010050002",3
"010050002",3
"010050002",3
"ninncnnn?" i
TYPE,DATE,SOI,N03
1.664137
1.664137
1.664137
1.664137
1. 664137
1.664137
1.664137
1. 664137
1. 664137
1.664137
1.664137
1.664137
1.664137
1.664137
1.664137
I
-35.60623
-85.60623
-85.60623
-85.60623
-85.60623
-85.60623
-85.60623
-85.60623
-85.60623
-85.60623
-85.60623
-85.60623
-85.60623
-85.60623
-85.60623
R, NH4, OCE, EC, CRUSTAL, 5ALT, DOW, H20_AIM, OC, N03
4,"STW",20031006,7.21,0,2.8,6.92,0.42,0.55,0
4,"STW",20031012,5.99,0,2.16,,,0.06,0.05,0.3
4,"STW",2 0031018,4.68,O, 1.75,5.58,0.55, 0.19,
4,"STW",2 003102 4,3.47,0,1.1,7.12,0.3 8,0.78,0
4,"STW",2 0031030,2.42,0,0.74,3.7,0.21,0.31,0
4,"STW",20031105,2.39,0,0.48,1.08,0.15,0.17,
4,"STW",2 0031111,1.43,0,0.34,2.91,0.2,0.28,0
4,"STW",2 0031117,2.76,0,0.69,1.8,0.2,0.3,0.0
4,"STW",2003112 3,3.41,0,1.21,3.23,0.5,0.24,0
4,"STW",2 003112 9,1.3,0,0.47,0.9,0.12,0.18,0,
4,"STW",2 00312 05,3.15,0,1.13,2.17,0.13,0.19,
4, "STW", 2 0031211,2.87,0, 1.03,0.69,0.13,0.09,
4,"STW",2 0031217, 1.88,0,0.76,0.3,0.13, 0.06,0
4,"STW",2 0031223,2.27,0,0.7,1.59,0.37,0.09,0
4,"STW",2 0031229,2.33,0,0.75,1.24,0.27,0.22,
"c;TTjrr 9r~iniiriirijl
ru n n n R1 n ? n n ft n
, S04_3 S , CI
,0.375,2 .!
61,2 .22, ,(
0,0.374,1
,0.317,1.:
.02,0.306.
0.02,0.20:
,0.238,0.;
1,0.25,0.!
,0.355, 1.;
0.3 62,0.4!
0,0.359,1
0,0.359,1
,0.375,0.1
.02,0.308.
0.06,0.32!
m n 11 r *
>
For Help, press F1
NUM
Speciated Monitor Data Variable Descriptions
Variable Description
Jd The ID is a unique name for each monitor in a particular location. (This is a
character variable.)
LAT Latitude at the monitor site in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are negative.
LONG Longitude at the monitor site in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g., United
States) are negative.
_TYPE STN or IMPROVE network data
DATE Date of daily average ambient data with YYYYMMDD format (This is a
numeric variable)
S04 Measured sulfate ion
N03R Estimated nitrate retained on FRM filter, using measured nitrate, hourly T and
RH
NH4 Measured NH4+ ion
Abt Associates Inc.
191
June 2010
-------
Daily PM Analysis: Details
OCB
EC
CRUSTAL
SALT
DON
H20_AIM
OC
N03
S04_3S
CRUSTAL_ALT
FRM_MASS
MEASURED_FM
RCFM
Al
Ca
Fe
Ti
Si
EPA_FLAG
USER FLAG
OC blank adjusted, STN uses constant blank value per sampler type,
IMPROVE uses backup filter methodology
Measured EC
Using IMPROVE algorithm, Crustal, aka "Fine Soil" = 2.2 * [Al] + 2.49 * [Si] +
1.63 x [Ca] + 2.42 x [Fe] + 1.94 x [Ti]
Estimated salt using CI (Salt=1,8*CI), where CI is elemental chloride
Degree of neutralization of S04 (0-0.375)
Calculated water using AIM and measured S04, adjusted N03 and measured
NH4
Measured OC
Measured nitrate ion
Sulfate value derived from S, as per IMPROVE protocol, i.e. S04_3S = 3*S.
Alternative Crustal calculation using measured Si, Fe, Ca, Ti (modified formula
without Al) = 3.73 x [Si] + 1.63 x [Ca] + 2.42 x [Fe] + 1.94 x [Ti]
FRM mass
STN or IMPROVE sampler measured fine mass (Teflon filter)
Reconstructed Fine Mass using IMPROVE protocol, RCFM = [Amm_Sulfate] +
[Amm_Nitrate] + [OCM] + [EC] + [Fine Soil]
Measured Al
Measured Ca
Measured Fe
Measured Ti
Measured Si
Flag to indicate data that EPA recommends to be removed from the species
fractions calculations. 0 = valid data, 1 or greater = data that has been flagged
and should be removed
Flag to indicate additional data that the user wants to remove from the species
fractions calculations. 0 = valid data, 1 or greater = data that has been flagged
and should be removed
Note: Some variables are supplied for QA purposes only and are either not used by MATS
or are calculated internally by MATS. For example, "OC" is not used by MATS (OCb is
used) and H20AIM is calculated internally. Character variables have names that begin
with an underscore (i.e., "_"), and the character values used can be kept with or without
quotes. (If a character variable has an embedded space, such as might occur with the name
of a location, then use quotes.)
7.3.2 PM2.5 Monitor Data Input
MATS uses the "unofficial" daily PM2.5 file for the calculation of species fractions files.
MATS uses the "official" daily PM2.5 file for point estimates.
7.3.2.1 Unofficial Daily PM2.5 Monitor Data Input
Monitor data should be in the form of a simple text file. The first row specifies the
frequency of the data (e.g., day). The second row presents comma-separated variable
names. The third row begins the data values. Below is an example of the monitor data file
Abt Associates Inc.
192
June 2010
-------
Daily PM Analysis: Details
format and descriptions of the variables in the file.
Format of Unofficial Daily PM2.5 Monitor Data
I
1 PM25-for-fractions-0206-v2.csv
-
WordPad
I File Edit View Insert Format
Help
~ & y a a #4
1
1
%
Courier New
10 v
Western
V
b / u j|3> m i m |E
[Day
_ID,_TYPE,LAT,LONG,DATE,PM25,EPA_FLAG,USER_FLAG
"010030010","FRH",30.498001046,-87.88141232,20020102,12.7,0,
"01003 0010","FRH",30.498001046,-87.881412 3 2,2 002 0105,12.5,0,
"01003 0010","FRH",30.498001046,-87.88141232,20020108,7.3,0,
"010030010","FRH",30.498001046,-87.881412 32,20020111,6.7,0,
"01003 0010","FRH",30.498001046,-87.88141232,20020114,5.7,0,
"010030010","FRH",30.498001046,-87.88141232,20020117,9.9,0,
"01003 0010","FRH",30.498001046,-87.88141232,20020120,9,0,
"01003 0010","FRH",30.498001046,-87.88141232,20020123,9.7,0,
"010030010","FRH",30.498001046,-87.881412 32,2002012 6, 10.2 ,0,
"01003 0010","FRH",30.498001046,-87.88141232,20020129,9.8,0,
"01003 0010", "FRH",30.498001046,-87.88141232,20020201,3.8,0,
"010030010","FRH",30.498001046,-87.88141232,20020204,5.4,0,
"01003 0010","FRH",30.498001046,-87.881412 3 2,2 002 02 07,6.2,0,
"010030010","FRH",30.498001046,-87.881412 32,20020210,10.2,0,
"01003 0010","FRH",30.498001046,-87.88141232,20020213,29,0,
"01003 0010","FRH",30.498001046,-87.881412 3 2,2 002 0216,13.3,0,
"ni nmnm n» "Fpm" 4QRnnin4fi -r^ rri41?t? ?nn?n?ig r 1 n
For Helpj press F1 MUM
Unofficial Daily PM2.5 Monitor Data Variable Descriptions
Variable
JD
_TYPE
LAT
LONG
DATE
PM25
EPA_FLAG
USER FLAG
Description
The ID is a unique name for each monitor in a particular location. (This is
a character variable.)
FRM, IMPROVE, or other "special" data - may include dummy sites.
(This is a character variable.)
Latitude at the monitor site in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are
negative.
Longitude at the monitor site in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g.,
United States) are negative.
Date of daily average ambient data with YYYYMMDD format (This is a
numeric variable)
Measured PM2.5 mass (ug/m3)
Flag to indicate data that EPA recommends to be removed from the
species fractions calculations. 0 = valid data, 1 = data that has been
flagged and should be removed
Flag to indicate additional data that the user wants to remove from the
species fractions calculations. 0 = valid data, 1 or greater = data that has
been flagged and should be removed.
Note: Character variables have names that begin with an underscore (i.e., and the
character values used can be kept with or without quotes. (If a character variable has an
Abt Associates Inc.
193
June 2010
-------
Daily PM Analysis: Details
embedded space, such as might occur with the name of a location, then use quotes.)
7.3.2.2 Official Daily PM2.5 Monitor Data Input
Monitor data should be in the form of a simple text file. The first row specifies the
frequency of the data (e.g., day). The second row presents comma-separated variable
names. The third row begins the data values. Below is an example of the monitor data file
format and descriptions of the variables in the file.
Format of Official Daily PM2.5 Monitor Data
i 24-hr-officiaL FRM 99 07 v2.csv WordPad
File Edit View Insert Format Help
~ #a #4 * sha,
%
Courier New
10
Western
B y U jjgl
|Day
_ID,
"01003
"01003
"01003
"01003
"01003
"01003
"01003
"01003
"01003
"01003
"01003
"01003
"01003
"01003
"01003
"01003
"01003
TYPE,
0010"
0010"
0010"
0010"
0010"
0010"
0010"
0010"
0010"
0010"
0010"
0010"
0010"
0010"
0010"
0010"
0010"
LAT, LONG, DATE, PH2 5,
"FRM",30.498001046,-87.
"FRH",30.498001046,-87.
"FRM",30.498001046,-87.
"FRH",30.498001046,-87.
"FRH",30.498001046,-87.
"FRH",30.498001046,-87 .
"FRH",30.498001046,-87.
"FRH",30.498001046,-87.
"FRH",30.498001046,-87.
"FRH",30.498001046,-87.
"FRH",30.498001046,-87 .
"FRH",30.498001046,-87.
"FRH",30.498001046,-87 .
"FRH",30.498001046,-87.
"FRH",30.498001046,-87.
"FRH",30.498001046,-87.
"FRH",3 0.49800104 6,-87 .
98_PERCENTILE, EPA_FLAG
88141232,20000116,5.9,0,
881412 32,2 0000119,11,0,0
881412 32,2 000012 2,8.6,0,
881412 32,2 000012 5,6.2,0,
881412 32,2 000012 8,7.6,0,
881412 32,2 0000131,9.1,0,
881412 32,2 00002 03,14,0,0
88141232,2 0000206,13.9,0
881412 32,2 00002 09,2 0.8,0
881412 32,2 0000310,14.3,0
881412 32,2 0000313,9.5,0,
881412 32,2 0000316,8.9,0,
881412 32,2 0000319,11.2,0
881412 32,2 00003 2 2,17.4,0
881412 32,2 00003 2 8,6.6,0,
881412 32,2 00003 31,2 4.1,0
881412 32,2 0000403,14.7,0
,COHPLETION CODE, STATE NAI
0,-9, "Alabama"
,-9,"Alabama","Baldwin"
0,-9,"Alabama"
0,-9,"Alabama"
0,-9,"Alabama"
0,-9,"Alabama"
,-9,"Alabama","Baldwin"
,0,-9,"Alabama'
,0,-9,"Alabama'
,0,-9,"Alabama'
0,-9,"Alabama",
0,-9,"Alabama",
,0,-9,"Alabama'
,0,-9,"Alabama'
0,-9,"Alabama",
,0,-9,"Alabama'
,0,-9,"Alabama'
Baldwin"
Baldwin"
Baldwin"
Baldwin"
Baldwin"
"Baldwin"
"Baldwin"
"Baldwin"
Baldwin"
Baldwin"
"Baldwin"
"Baldwin"
Baldwin"
"Baldwin"
"Baldwin"
For Help, press F1
NUM
Official Daily PM2.5 Monitor Data Variable Descriptions
Variable
JD
_TYPE
MONTIOR_LAT
MONITOR LONG
DATE
PM25
98 PERCENTILE
Description
The ID is a unique name for each monitor in a particular location. The default
value is the AIRS ID. (This is a character variable.)
FRM data
Latitude at the monitor site in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are negative.
Longitude at the monitor site in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g., United
States) are negative.
Date of daily average ambient data with YYYYMMDD format (This is a
numeric variable)
Measured PM2.5 mass (ug/m3)
Official 98th percentile values for each site for each year are identified by a
flag = "1"
Abt Associates Inc.
194
June 2010
-------
Daily PM Analysis: Details
EPA_FLAG Official flags to indicate data that has been removed from design value
calculations. 0 = valid data, 1 = data that has been flagged and removed
USER_FLAG Flag to indicate additional data that the user wants to remove from design
value calculations. 0 = valid data, 1 = data that has been flagged and should
be removed.
COMPLETION_CODOfficial design value completion codes (1, 2, 3, 4, 5, and -9). Codes are valid
E for the end year of each 3 year design value period.
_STATE_NAME State name. (This is a character variable.)
_COUNTY_NAME County name. (This is a character variable.)
Note: Character variables have names that begin with an underscore (i.e., and the
character values used can be kept with or without quotes. (If a character variable has an
embedded space, such as might occur with the name of a location, then use quotes.)
7.3.3 Model Data Input
Model data should be in the form of a simple text file. The first row specifies the
frequency of the data (e.g., day). The second row presents comma-separated variable
names. The third row begins the data values. Below is an example of the model data file
format and descriptions of the variables in the file. Note that there is both a base year and
a future year model file. The format for both is the same.
Note that you can load in either daily model data or quarterly model data.
Model Data
® Daily model data input O Quarterly peak model data input
Baseline File [CAProgram Files Vtot Associates'1',MATS\S ampleData\2002cc_EUS_F'M25_sub.csv
Forecast File IC:\Proqram Files\Abt Associates\MAT SVElampleD ata'\2020cc_E U S_PM 25_sub. csv
Format of Model Data
Abt Associates Inc.
195
June 2010
-------
Daily PM Analysis: Details
2002cc_EUS_PM25_sub.csv - WordPad
File Edit View Insert Format Help
b x u ^ ^ ^ -E
|day
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
1500
ID, TYPE,
,32
,32
,32
,32
,32
,32
,32
,32
,32
,32
,32
,32
,32
,32
,32
,32
,32
,32
,32
,32
,32
,32
,32
,32
,32
,32
,32
LAT
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
041036,
LONG ,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
-88.664718,
DATE ,
20020101,
20020102,
20020103,
20020104,
20020105,
20020106,
20020107,
20020108,
20020109,
20020110,
20020111,
20020112,
20020113,
20020114,
20020115,
20020116,
20020117,
20020118,
20020119,
20020120,
20020121,
20020122,
20020123,
20020124,
20020125,
20020126,
20020127,
Crustal,
2.8938
9910
1356
2023
3665
4842
7300
8170
8733
9303
6944
4354
6101
4628
1597
6949
552 6
8016
2738
8043
9837
0872
6651
6103
2452
0805
NH4 ,
0.8440,
0064,
3274,
4508,
1407,
1455,
1185,
1678,
1544,
7045,
8669,
4571,
1805,
2882,
6886,
6052,
4553,
72 67,
4397,
9561,
8742,
0715,
2512,
3232,
6444,
1094,
0487,
S04 ,
1.1525,
2453,
5503,
2391,
6652,
3141,
0372,
0973,
1099,
2925,
8649,
6251,
9503,
0504,
8693,
9658,
4824,
6180,
1787,
8673,
2578,
3561,
8760,
2327,
9694,
0850,
0349,
EC ,
,3428,
,3024,
,3894,
,5332,
,4050,
,0818,
,2976,
,4232,
,6474,
,3994,
,4779,
,3200,
,5070,
,5513,
,4941,
,3923,
, 6301,
,3462,
,2426,
,3597,
,3345,
,4460,
,1455,
, 1413,
,2579,
,5502,
4185,
8583,
6868,
4481,
8896,
1343,
5331,
6045,
5588,
9135,
7046,
7874,
8407,
8685,
2697,
8483,
9499,
4840,
1327,
0177,
3031,
7230,
0104,
0067,
9901,
2805,
2156,
OC ,
,3510,
,1155,
,2507,
,0241,
,2245,
,4239,
,2708,
,7906,
,5978,
,2099,
,2104,
,7154,
,7200,
,8608,
,4518,
, 1626,
,7062,
,8798,
,2229,
, 6218,
,5093,
,7432,
,5435,
,4728,
,9955,
,5656,
,6647,
PM25 ,
8.3861,
8.8275,
11.6806,
13 .4746,
11.4404,
1.7175,
9.3578,
10.4348,
11.7966,
8.2250,
9.5399,
6.8945,
12.7001,
13.0609,
9.7304,
8.3897,
15.0897,
7.4988,
4.8955,
9.1079,
7.7430,
11.3870,
2.6622,
2.9357,
6.3848,
12 .4952,
13.1366,
4677
3373
4237
9857
5842
1237
4677
8712
7997
1342
7523
6504
0805
0743
0514
7299
102 6
6300
3462
6233
7217
7273
6255
8523
2 644
8330
6917
-or Helpj press F1
Model Data Variable Descriptions
Variable Description
_ID The ID is a unique name for each monitor in a particular location. The default
value is the column identifier multiplied by 1000 plus the row. (This is a
character variable.)
_TYPE Leave blank
LAT Latitude at the grid cell centroid in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are negative.
LONG Longitude at the grid cell centroid in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g., United
States) are negative.
DATE Date of daily average model value with YYYYMMDD format (This is a numeric
variable)
CM Coarse PM mass (ug/m3)
Crustal Crustal PM2.5 mass
NH4 Ammonium mass
S04 Sulfate PM
EC Elemental carbon
N03 Nitrate PM
OC Organic mass PM
PM25 PM2.5 mass
SALT Salt (Optional)
Note:
The "PM25" mass variable is used to calculate PM2.5 model "gradients" for gradient
Abt Associates Inc.
196
June 2010
-------
Daily PM Analysis: Details
adjusted spatial fields. It is up to the user to provide a modeled PM2.5 mass
concentration using an appropriate definition of modeled PM2.5 mass.
Character variables have names that begin with an underscore (i.e., "_"), and the
character values used can be kept with or without quotes. (If a character variable has an
embedded space, such as might occur with the name of a location, then use quotes.)
7.4 Species Fractions Calculation Options
The Species Fractions Calculation Options has two main sections. One involving
speciated monitor data (e.g., STN and IMPROVE monitors) and the other total PM2.5
monitor data (FRM and IMPROVE). For each type of data you can specify the years of
interest, whether you want to delete certain data, and the minimum amount of data for a
monitor to be considered "valid" (and thus included in the calculations). In the next
sections, we describe each of these three options in more detail. Note that these options
apply more or less in the same way for both speciated monitor data and total PM2.5
monitor data. Also note that these options are no longer relevant if you have loaded a
species fractions file.
IMPROVE-STN Monitor Data. The speciation data from STN and IMPROVE
monitors are interpolated by MATS in order to provide species data for any point in a
modeling domain. The interpolated species data is used to calculate species fractions at
FRM monitors. Note that you do not need to have values for all species for a monitor to
be considered valid, as each species is considered individually. However, the
"EPA Flag" variable in the default "species for fractions" file has been set so that all
monitor days that do not have complete species data are not used in the calculations (flag
= 1). If the user wants to use the incomplete species data, the flag can changed to "0".
PM2.5 Monitor Data. The total PM2.5 data from FRM are used by MATS to calculate
species fractions (in conjunction with the interpolated speciation data from STN and
IMPROVE monitors).
Abt Associates Inc.
197
June 2010
-------
Daily PM Analysis: Details
I Daily PM Analysis
Qnor*ioc Frartinnc Palr*i il^tinn Ontinnc
O 1 1 uvllUI IO ^ulUUIallUI 1 WpilUI IO
factions Options
Species Fractions • Advanced
PM2.5 Calculation Options
Model Data Options
Final Output and Check
IMPROVE-STN Monitor Data
Monitor Data Years
Start Year End Year
ISHHEI [20M-
Delete Specified Data Values
p EPA-specified deletions from monitor data
f~ User-specified deletions from monitor data
Minimum Data Requirements
Minimum number of valid days per quarter
Minimum number of valid quarters per valid year
Minimum number of valid years required for valid monitor
PM2.5 Monitor Data
Monitor Data Years
Start Year End Year
12002 3 [2004 3
Delete Specified Data Values
EPA-specified deletions from monitor data
User-specified deletions from monitor data
Minimum Data Requirements
Minimum number of valid days per quarter
Minimum number of valid quarters per valid year (point calculations) [
Minimum number of valid years required for valid monitor
ITU
~~i±l
11±]
33
"^3
< Back
Next >
Cancel
7.4.1
Monitor Data Years
Using the Start Year and End Year drop-down menu options, you can choose more than
one year of speciated data for the calculation of species fractions. The default approach in
MATS is to use three years of data.
Monitor Data Years
Start Year End Year
12002
~
2003
H -j
2004
2005
2006
a
2004
;letions from monitor data
MATS handles multiple years of data by calculating averages for each species by quarter
and year. MATS then averages the quarterly values across the years (e.g., average quarter
1 values of S04 across two years to get a single "quarter 1" estimate). For example, if for
Abt Associates Inc.
198
June 2010
-------
Daily PM Analysis: Details
2002-2004 we have S04 averages of peak values in quarter 1 of 8.2 ug/m3 (for 2002), 6.2
ug/m3 (for 2003), and 3.6 ug/m3 (for 2004), then the average across the years would equal
6.0 (=[8.2+6.2+3.6]/3).
After completing this step, MATS will have four quarterly estimates for each species at
each speciated monitor. These quarterly values are then ready to be interpolated to FRM
sites.
7.4.2 Delete Specifed Data Values
The default is to delete the observations specified by EPA. (That is, these observations are
excluded from a particular analysis, while they of course remain in the database.) As
described in the Data Input section, valid data are given a value of "0" and observations
that are deleted are given a value of "1" to "10", as follows:
0. Data is OK to use
1. Important species is missing (e.g. no3, so4, oc, etc)
2. Constructed mass < 30% total mass
3. Constructed mass > 2 times total mass
4. Fire event
5. total mass < so4
6. total mass < crustal
7. OC is outside QA criteria
8. Soil is outside QA criteria
9. Both OC and soil are outside QA criteria
10. Regional concurrence on exceptional event
There is also an option for the user to flag data, using the same convention of "0" for valid
data and "1" to "10" for data marked for deletion. If both the EPA-specified and User-
specified flags are checked, then MATS deletes any observations that are marked for
deletion by either the EPA or the user. This makes it easy for the user to flag additional
data for removal from the calculations (without deleting the actual record from the ambient
data file).
Delete Specified Data Values
P" EPA-specified deletions from monitor data
~~ User-specified deletions from monitor data
7.4.3 Minimum Data Requirements
There are three sets of minimum data requirements:
Minimum number of valid days per quarter. This is the minimum number of
site-days per valid quarter. The default is 11 days, which corresponds to > 75%
completeness for monitors on a 1 in 6 day schedule. This is a minimum number of
Abt Associates Inc.
199
June 2010
-------
Daily PM Analysis: Details
samples that is routinely used in calculations of quarterly average concentrations.
Minimum number of valid quarters per valid year. This is the number of valid
quarters of data required for a given year. The default value is 1 quarter. If the value is
set = 2, then there will need to be 2 quarters of valid data for a given year to be valid.
Minimum number of valid years required for valid monitor. This is the number of
valid years that are needed in order for a particular monitor's data to be considered valid.
The default is 1 for IMPROVE-STN monitor data and the range is 1-4.
Minimum Data Requirements
Minimum number of valid days per quarter | 11
Minimum number of valid quarters per valid year i—m
Minimum number of valid years required for valid monitor
Example 1: Minimum Days = 11, Minimum Quarters = 1, Minimum Years = 1
Consider the default assumptions and the following number of observations from three
monitors:
Monitor 1
Monitor 2
Monitor 3
Year
Quarter
# Obs.
# Obs.
# Obs.
2002
1
8
11
11
2
6
13
8
3
11
4
12
4
5
12
8
2003
1
11
10
9
2
9
11
12
3
12
13
10
4
6
12
9
2004
1
6
15
12
2
7
11
10
3
12
12
10
4
13
9
12
With the default 11 minimum days required, MATS would then consider the highlighted
quarters (above) as valid. MATS would then use all of the highlighted quarters as all of the
following years are valid.
Monitor 1 Monitor 2 Monitor 3
Year Avg (ug/m3) Avg (ug/m3) Avg (ug/m3)
2002 V V V
2003 V V V
2004 V V V
Example 2: Minimum Days = 11, Minimum Quarters = 2, Minimum Years = 2
Abt Associates Inc.
200
June 2010
-------
Daily PM Analysis: Details
Consider the data from three monitors in the previous examples:
Minimum Days = 11, Minimum Quarters = 2, Minimum Years = 2
Monitor 1
Monitor 2
Monitor 3
Year
Quarter
# Obs.
# Obs.
# Obs.
2002
1
8
11
11
2
6
13
8
3
11
4
12
4
5
12
8
2003
1
11
10
9
2
9
11
12
3
12
13
10
4
6
12
9
2004
1
6
15
12
2
7
11
10
3
12
12
10
4
13
9
12
With the default 11 minimum days required, MATS would then consider the highlighted
quarters (above) as valid. MATS would then use only a subset of the highlighted quarters,
as not all of the years are valid. Monitor 1 would not be used at all, and data from Monitor
3 would only be used from 2002 and 2004.
Monitor 1 Monitor 2 Monitor 3
Year Avg (ug/m3) Avg (ug/m3) Avg (ug/m3)
2002 - V V
2003 - V
2004 - V V
7.5 Species Fractions Calculation Options - Advanced
The Species Fractions Calculation Options - Advanced screen allows you to make
relatively advanced choices for your analysis. Generally speaking, the default options
settings are consistent with the EPA modeling guidance document. The first set of options
lets you specify which monitor data you want to use. The second set of options allows you
to specify the interpolation weighting that you want to use and whether the interpolation
involves a maximum distance or not. The third set of options involves choices regarding
ammonium, blank mass, and organic carbon.
Abt Associates Inc.
201
June 2010
-------
Daily PM Analysis: Details
Daily PM Analysis
^^Ispecies Fractions - Advanced
¦ PM2.5 Calculation Options
¦ Model Data Options
I Final Output and Check
Species Fractions Calculation Options - Advanced
Using Monitor Data to Calculate Species Fractions
IMPROVE-STN Monitor Data
® iUse top X percent of daily monitor days! | 10-$~|
Use all daily monitor values greater than i ri
fixed amount (uq/m3)
Minimum number of days required above fixed amount | 1 ~~t\
O Use top X number of daily monitor days | 25-^-j
PM2.5 Monitor Data
<•) Use top X percent of daily monitor days | 10~ri
Use all daily monitor values greater than i Q-ri
~ fixed amount (uq/m3) ' ^
Minimum number of days required above fixed amount j 1 ~r|
Q Use top X number of daily monitor days | 25-—¦]
Interpolation Options
PM2.5 |~ 90000j< Crustal | Inverse Distance Squared V ~ 190000j< E:».
S04 | Inverse Distance Squared V ~] [90000j« DON
N03 | Inverse Distance Squared V ~] 190000 *}~7"| »\ OC
EC | Inverse Distance Squared V ~ ] |90000 **| ^H4
Salt | Inverse Distance Squared V ~ ] 190000 *|~$~| i»|
Miscellaneous Options
Ammonium
® Use DON values
Use measured ammonium
NH4 percentage evaporating (0-100)
| Inverse Distance Squared V ~ 1 |90000 »|
| Inverse Distance Squared V ~ [ |90000 »|
nverse Distance Squared J l »: 0 H I
03
Default Blank Mass
Default Blank Mass
Organic Carbon
Organic carbon mass balance floor f
Organic carbon mass balance ceiling
0.5±|
~r±l
0.G^j
< Back
Next >
Cancel
7.5.1 Using Monitor Data to Calculate Species Fractions
The Using Monitor Data to Calculate Species Fractions panel allows you to choose how
you will choose IMPROVE and STN speciated monitor data and ("unofficial") PM2.5
monitor data to calculate quarterly peak values. There are three options:
Use Top X Percent of Daily Monitor Days. MATS chooses the top "X" percent of
days that you specify and averages them.
Use All Daily Monitor Values Greater than Fixed Amount (ug/m3). MATS chooses
all monitor values greater than or equal to a fixed amount that you specify and averages
them. Note that you also need to specify a minimum number of days greater than or
Abt Associates Inc.
202
June 2010
-------
Daily PM Analysis: Details
equal to your specified amount. If there are an insufficient number of days then MATS
drops the data for that particular quarter.
Use Top X Number of Daily Monitor Days. MATS chooses the top "X" number of
days that you specify and averages them.
These three options work in essentially the same way for IMPROVE and STN speciated
monitor data and for ("unofficial") PM2.5 monitor data. Below are some examples of how
these options work.
Note that the IMPROVE and STN speciated monitor data uses the "measured FM"
variable to identify peak days. Note also that it is possible that the option used to identify
peak days (e.g., Use Top X Percent of Daily Monitor Days) does not give a unique set of
days because a number of days may have the same value (see Example 1, below).
Using Monitor Data to Calculate Species Fractions
IMPROVE-STN Monitor Data
© Use top X percent of daily monitor days!
Use all daily monitor values greater than
fixed amount (uq/m3)
Minimum number of days required above fixed amount
Use top X number of daily monitor days
PM2.5 Monitor Data
• Use top X percent of daily monitor days
Use all daily monitor values greater than
fixed amount (uq/rn3)
Minimum number of days required above fixed amount
Use top X number of daily monitor days
Tog
~°I]
25^
log
"og
1±I
"25 m
Example 1: Use Top 10 Percent of Daily Monitor Days
Assume the default of using the top 10 percent of days. Let's say there are 27 days in the
first quarter (January-March) at a particular monitor for a given year, and 24, 25, and 32
valid days of data in quarters 2,3, and 4. MATS will then calculate the number of peak
days in each quarter as follows:
Year 2002, Ql: round(0.10*27) = round(2.7) = 3
Year 2002, Q2: round(0.10*24) = round(2.4) = 2
Year 2002, Q3: round(0.10*25) = round(2.5) = 3
Year 2002, Q4: round(0.10*32) = round(3.2) = 3.
In the sample data for quarter 1 (below), there are not three unique peak days because the
Abt Associates Inc.
203
June 2010
-------
Daily PM Analysis: Details
third and fourth highest days (highlighted in yellow) have the same value. In this case,
MATS uses the top 4 days.
Sample Data for Quarter 1 - Highlighted Top 10 Percent
Day
measured FM
1
13.9
2
13.8
3
13.6
4
13.6
5
13.0
6
12.6
7
12.6
8
12.4
9
12.1
10
11.7
11
11.4
12
10.4
13
10.2
14
10.1
15
9.6
16
9.5
17
9.4
18
9.0
19
8.9
20
8.4
21
8.2
22
8.2
23
7.6
24
7.4
25
7.3
26
6.6
27
6.6
Example 2: Use All Daily Monitor Values Greater than 13 ug/m3
With the same sample data and with the choice to use monitor values greater than 13 ug/m3
, MATS will select the top five days. Note that if you had selected values greater than 14
ug/m3, MATS would not have any data and would have dropped this quarter.
Sample Data for Quarter 1 - Highlighted Greater than or Equal to 13 ug/m3
Day
measured FM
1
13.9
2
13.8
3
13.6
4
13.6
5
13.0
6
12.6
7
12.6
8
12.4
9
12.1
10
11.7
11
11.4
12
10.4
Abt Associates Inc.
204
June 2010
-------
Daily PM Analysis: Details
13
10.2
14
10.1
15
9.6
16
9.5
17
9.4
18
9.0
19
8.9
20
8.4
21
8.2
22
8.2
23
7.6
24
7.4
25
7.3
26
6.6
27
6.6
Example 3: Use Top 3 Daily Monitor Days
With the same sample data and with the choice to use the top three monitor days, MATS
would select the top four days, because the third and fourth days have the same value. This
is the same result as in Example 1 (above).
7.5.2 Interpolation Options for Species Fractions Calculation
The Interpolation Options panel allows you to choose how you will interpolate, or
combine, the values from different monitors. One approach is to use Inverse Distance
Weights. This means that the weight given to any particular monitor is inversely
proportional to its distance from the point of interest. A second approach is Inverse
Distance Squared Weights, which means that the weights are inversely proportional to the
square of the distance. And the third approach is Equal Weighting of Monitors. The
default approach for PM is Inverse Distance Squared Weights.
Interpolation Options
IInverse Distance Squared ^ w
Crustal [inverse Distance Squared •* |900000 «
HHH
| Inverse Distance Squared ^ T [
|300000 « »
DON | Inverse Distance Squared 19000~ 0 ««
Equal Weighting of Monitors
Inverse Distance Weights
)000 * »
OC [inverse Distance Squared ^ ~ | |900000 «
(inverse Distance Squared Vv'eiqhts:
)000 « »
NH A
| Inverse Distance Squared ^ T [
900000
When interpolating monitor values, MATS allows you to identify the monitors you want to
use based on their distance away from the point of interest (e.g., the center of a grid cell).
The first step in the interpolation process is to identify the monitors that are nearby, or
neighbors, for each point of interest. The next step is to determine the distance (in
kilometers) from the nearby monitors to the point of interest.
The default approach is to include all valid monitors (i.e., those that satisfy the three
criteria in the Species Fractions Calculation Options panel), regardless of distance. If you
Abt Associates Inc.
205
June 2010
-------
Daily PM Analysis: Details
want to limit the use of monitors based on distance, type in the distance you want to use
(e.g., 100) next to the pollutant of interest.
Interpolation Options
PM2.5 [inverse Distance Squared \ T
Crustal J Inverse Distance Squared \ T | |900000 « , >*
SO4 [inverse Distance Squared | iUlll «
W-
DON Inverse Distance Squared \ T | J900000 « , >*
NOx |Inverse Distance Squared^ ~ | |900000 «
H-
OC Inverse Distance Squared \ T [900000 « „ »
EC [inverse Distance Squared \ » | |900000 «
NH4
Salt [inverse Distance Squared \ T [ [900000 « ,, »
You can also change the number using the arrows. The double arrow on the right
increases the number in units of 100:
44
w
and the double arrow on the left decreases the number in units of 100. The upper arrow
increases the number in single digits:
205 *
and the lower arrow reduces the number in single digits.
Note that a distance of one hundred (100) kilometers means that any monitors further than
100 kilometers can no longer be used in the interpolation. If a point of interest has no
monitors within the specified distance, then no value is calculated.
7.5.3 Miscellaneous Options
The Miscellaneous Options panel lets you make choices regarding:
Ammonium. This allows you to specify whether MATS uses degree of neutralization
(DON) values to calculate ammonium (NH4) or whether it uses measured ammonium in
conjunction with an assumption about the percentage of NH4 that evaporates. The
default option is to use DON values. If you want to use measured ammonium, you need
to click the button and choose a NH4 percentage evaporating (e.g., 50). The default is
"0", which assumes that no ammonium evaporates from the FRM filters. The
calculations underlying the default and alternative ammonium calculations are discussed
in detail in the section on species fractions calculations.
Default Blank Mass. The Default Blank Mass option simply allows you to set default
blank mass to the desired level. The default is 0.5. You can type desired value, or use
the arrows to increase or decrease the value.
Organic Carbon. This allows you to set the "floor" and the "ceiling" for the mass
balance calculation for organic carbon. The calculations involved are discussed in detail
Abt Associates Inc.
206
June 2010
-------
Daily PM Analysis: Details
in the section on species fractions calculations.
Miscellaneous Options
Ammonium
© Use DON values
O Use measured ammonium
NH4 percentage evaporating (0-100)
Default Blank Mass
Default Blank Mass | 0.5~^-|
Organic Carbon
Organic carbon mass balance floor | 1 vj
Organic carbon mass balance ceiling 0.8~^-|
7.5.4 Internal Precision of the Calculations
All calculations in MATS are carried out with single precision. In addition, most output
files by default generate outputs only up to 3 digits after the decimal. Therefore, the base
year DV (b_pm25_d_q_DV) may differ slightly from the sum of the component species of
the quarterly PM2.5 MATS output files. This issue may be remedied by increasing
precision of the species and species fractions calculations from the default 3 significant
digits to 7 (or more) significant digits by modifying the MATS.ini file as follows: set
species_calc_precision=7 and species_fraction_precision=7. Please note that the future
year species always add up to the future year DV. However, increases in species fractions
precision may result in very small changes in future DV due to the dependence of the
future concentrations on the base year concentrations.
7.6 PM2.5 Calculation Options
The PM2.5 Calculation Options window allows you to specify the particular years of
monitor data that you want to use from the input file you specified in the Data Input
section. You can also specify the following:
Valid FRM Monitors. You can specify the minimum number of design values (the
default is 1) and whether you want to make sure that particular design values have to be
used in the calculations.
NH4 Future Calculation. You can also specify how you want to forecast NH4 values.
The default approach is to use baseline DON values, and the alternative is to use baseline
NH4 and a RRF value for NH4. (These calculations are described in detail here.)
Abt Associates Inc.
207
June 2010
-------
Daily PM Analysis: Details
Daily PM Analysis
Choose Desired Output
¦ Output Choices-Advanced
¦ Data Input
¦ Species Fractions Options
¦ Species Fractions - Advanced
PM2.5 Calculation Options
¦ Model Data Options
¦ Final Output and Check
PM2.5 Calculation Options
PM2.5 Monitor Data Years
Start Year 12000 * EndYear 12004 »
Valid FRM Monitors
Minimum Number of Design Value Periods fl ^
Required Design Value Periods |None selected
-
_NH4 future calculation
f* Calculate future year NH4 using base year (constant) DON values
r Calculate future year NH4 using base year NH4 and the NH4 RRF
7.6.1 PM2.5 Monitor Data Years
Using the Start Year and End Year drop-down menu options, you can choose more than
one year of official PM monitor data for the calculation of future PM2.5 values. The
default approach in MATS is to use five years of data.
PM2.5 Monitor Data Years
* Officiej
Minimum d.
[2000
EndYear 2004
1999
2000
2001
2002
5
2003
2004
2005
llll
2006
7.6.2 Valid FRM Monitors
By default, MATS assumes that there only needs to be one design value for a monitor to be
considered valid. In addition, MATS assumes that no particular design value is required,
so different monitors with different years of data could be used. For example, if you
specify the start year and end year as 2000 and 2004 (giving potential design values of
2002, 2003, and 2004), then one monitor could have data for, say, 2002 and another
monitor data for 2003, and both monitors would be used.
Abt Associates Inc.
208
June 2010
-------
Daily PM Analysis: Details
Valid FRM Monitors
Minimum Number of Design Value Periods F 3
Required Design Value Periods [None selected T
7.6.3 NH4 Future Calculation
As described in the section on Forecasted Species Calculation. MATS can forecast NH4
using two different approaches. The default approach is to use base year DON values.
NH4 future calculation
(* Calculate future year MH4 using base year (constant) DON values
C Calculate future year NH4 using base year NH4 and the NH4 RRF
7.7 Model Data Options
The Model Data Options section allows you to specify:
Temporal Adjustment at Monitor. This option specifies how many model grid cells to
use in the calculation of RRFs for point estimates and for spatial estimates. Using the
drop-down menu, you can choose lxl, 3x3, 5,x5, and 7x7. (The default for a 12
kilometer by 12 kilometer grid is to use a lxl set of grid cells. The choice of grid size is
discussed in the following EPA guidance:
http://www.epa.gov/scram001/guidance/guide/final-03-pm-rh-guidance.pdn ForPM
analyses, MATS calculates mean concentrations across the grid cell array (as compared
to maximum concentrations used for ozone analyses).
Advanced Options: RRF Model Values Used. This option allows you to choose three
different ways to calculate quarterly peak values: (1) the top X percent of daily model
days, (2) all daily model values greater than or equal to a specified amount, and (3) the
top X number of model days. (These three options are described in detail in Species
Fractions Calculation Options - Advanced section.)
Max Distance to Domain. This is the maximum distance from a given monitor to the
nearest model grid cell (measured in kilometers). If a monitor is further than the
specified maximum distance from the center of any grid cell, then MATS will not
calculate results for that monitor or use the monitor in any calculations (this effectively
eliminates monitors from outside the specified domain from the being included in the
output files).
Abt Associates Inc.
209
June 2010
-------
Daily PM Analysis: Details
Daily PM Analysis
Choose Desired Output
¦ Output Choices-Advanced
¦ Data Input
¦ Species Fractions Options
¦ Species Fractions - Advanced
¦ PM 2.5 Calculation 0 ptions
Model Data Options
Final Output and Check
Model Data Options
7.8 Final Check
The Final Check window verifies the selections that you have made.
Abt Associates Inc.
210
June 2010
-------
Daily PM Analysis: Details
Daily PM Analysis
Choose Desired Output
¦ Output Choices-Advanced
¦ Data Input
¦ Species Fractions Options
¦ Species Fractions - Advanced
¦ PM 2.5 Calculation 0 ptions
¦ Model Data Options
Final Output and Check
Final Output Choices and Verification
Verify inputs
| Press here
Checking...
Check OK. Press the finish button to continue..
Save Scenario
Save Scenario St Run
Click the button Press here to verify your selections. If there are any errors, MATS will
present a message letting you know. For example, if the path to a model file is invalid —
perhaps you misspelled the file name — you would get the following error:
Verify inputs
E yQur selections...
Checking...
Please verify the Model Data Baseline File setting.
Check OK. Press the finish button to continue..
After making the necessary correction, click the button Press here to verify your
selections. Then click the Finish button.
Abt Associates Inc.
211
June 2010
-------
Daily PM Analysis: Details
- Verify inputs
[i Press here to.ven^our seiections... i|
Checking...
Check OK. Press the finish button to continue..
7.8.1 Running MATS in Batch Mode
The Save Scenario button will save the scenario as a configuration file (.cfg file). The
"*.cfg" file will be saved in the .\MATS\output directory. Several .cfg files can be created
with the MATS interface and run later in batch mode. To do this, edit the default batch file
located in the .\MATS directory. The file "batchmats.bat" should be edited with a text
editor to point to the name and location of the .cfg files that will be run in batch mode.
& U It raE dit - 3 2 - [C: \Doc ume nts a n d Setti ngs\btimi n\Deskto p\M ATS - 5 -0 8\batc hmats. bat] |"^~| [~] | X J
& File Edit Search Project View Format Column Macro Advanced Window Help _ n1 X
D Or eJ1 S a H A I Wt | H | r X ir=i ® ^mT^IfPgo tp 0 HM
batchmats.bat
start /wait MATS.exe "c:\Program Files\Abt Associates\MATS\\output\default.cfg" ^
start /wait MATS.exe "c:\Prograiri Files\Abt Associates\HATS\\output\default2.cfg"
.A I jf*
For Help, press F1 Ln 1, Col. 1, CW DOS Mod: 5/23/2008 5:02:49PM File Size: 163 INS | ^
After editing the batchmats.bat file, simply run the .bat file. MATS will start and run in
the background.
Abt Associates Inc.
212
June 2010
-------
Ozone Analysis: Quick Start Tutorial
8 Ozone Analysis: Quick Start Tutorial
In this tutorial you will forecast ozone design values at monitors in the Eastern United
States. The steps in this analysis are as follows:
Step 1. Start MATS. Start the MATS program and choose to do an Ozone analysis.
Step 2. Output Choice. Choose the output to generate. In this example, you will
forecast ozone levels at monitor locations.
Step 3. Data Input. Choose the data files for input to MATS.
Step 4. Filtering & Interpolation. Choose the particular years of data and monitors to use
in this analysis.
Step 5. RRF & Spatial Gradient. Specify how to generate the relative response factors
fRRFs) used in the forecasts.
Step 6. Final Check. Verify the choices you have made.
Step 7. Load & Map Output. Load your output and prepare maps of your forecasts.
Step 8. View & Export Output. Examine the data in a table format and export these
data.
Each step is explained below. Additional details are provided in the section Ozone
Analysis: Details.
8.1 Stepl. Start MATS
Double-click on the MATS icon on your desktop, and the following window will appear:
Abt Associates Inc.
213
June 2010
-------
Ozone Analysis: Quick Start Tutorial
Umats
Help T
Start Map View Output Navigator
Ozone Analysis
Visibility Analysis
Annual PM Analysis
Daily PM Analysis
Stop Info
Click the Ozone Analysis button on the main MATS window. This will bring up the
Configuration Management window.
Configuration Management
(* plreate New Configuration!
C Open Existing Configuration
Go
Cancel
A Configuration allows you to keep track of the choices that you make when using MATS.
For example, after generating results in MATS, you can go back, change one of your
choices, rerun your analysis, and then see the impact of this change without having to enter
in all of your other choices. For this example, we will start with a New Configuration.
Choose Create New Configuration and click the Go button. This will bring up the
Choose Desired Output window.
8.2 Step 2. Output Choice
The Choose Desired Output window allows you to choose the output that you would like
Abt Associates Inc.
214
June 2010
-------
Ozone Analysis: Quick Start Tutorial
to generate. MATS allows you to forecast Point Estimates (at ambient monitors) or to
generate a Spatial Field of either Baseline values or Forecast values.
In the Scenario Name box type " Tutorial 03" - this will be used to keep track of where
your results are stored and the variable names used in your results files. Leave the box
checked next to Temporallv-adiust ozone levels at monitors. MATS will create forecasts
for each monitor in the monitor file.
Check the box next to Automatically extract all selected output files. MATS will create a
separate folder called "Tutorial 03" in the MATS "Output" folder, and then export .CSV
files with the results of your analysis. Alternatively, you can export the results from the
Output Navigator, but checking this box is a little easier.
Choose Desired Output
Scenario Name : |Tutorial 03
Point Estimates
Forecast
p T emporally-adjust ozone levels at monitors.
Spatial Field
Baseline
Interpolate monitor data to spatial field
Interpolate gradient-adjusted monitor data to spatial field.
Forecast
Interpolate monitor data to spatial field. Temporally adjust ozone levels.
Interpolate gradient-adjusted monitor data to spatial field. Temporally adjust.
Actions on run completion
V/ Automatically extract all selected output filed
< Back
Next >
Cancel
When your window looks like the window above, click Next. This will bring you to the
Data Input window.
8.3 Step 3. Data Input
The Data Input window allows you to choose the monitor data and the model data that
you want to use. As discussed in more detail in the following chapter (see RRF Setup).
MATS calculates the ratio of the base and future year model data to calculate a relative
response factor (RRF). MATS then multiplies the design value from the monitor data with
the RRF to calculate a future-year design value.
MATS currently comes loaded with ozone design values for the period from 1997-2005
Abt Associates Inc.
215
June 2010
-------
Ozone Analysis: Quick Start Tutorial
(1997-1999, 1998-2000, 1999-2001, 2000-2002, 2001-2003, 2002-2004, and 2003-2005);
and it comes loaded with example ozone model data for 2001 and 2015. These are the
files needed to calculate the Point Estimates and Spatial Fields listed in the Desired Output
window.
Use the default settings in the Data Input window. The window should look like the
following:
B
H? Desired output
Data Input
I Filtering/Interpolation
Data Input
¦ h h h /b patial Li radient
ZM Final Check
Monitor Data
~ zone Data
| rnpleD ata\DZ0 N E_MAT 5_input_9907-v2. csv ¦ ¦ ¦"}
Model Data
Baseline File
Forecast File
Using Model Data
T ernporal adjustment at monitor [
15 \S arnpleP ata\ozone_model_data_2001. csv ¦ • ¦ |
IS \5 ampleD ata\ozone_model_data_2015. csv ¦ ¦ ¦ |
3x3 ~ I [Maximum ~ |
< Back
Next >
Cancel
Note that MATS gives you the option to use model data in different ways when calculating
forecasts at each monitor. The user can choose to use the model results from the single
grid cell that contains the monitor or select a grid cell array of 3x3, 5x5, or 7x7 model cells
around each monitor. The example model output dataset contained in MATS is at 12km
resolution. Therefore, for this example, a 3x3 grid cell array should be used (see section
3.2 of the modeling guidance). The default for ozone analysis is to choose the maximum
value each day in the array for the calculation. This is described in more detail in the
Using Model Data section of the Ozone Analysis: Details chapter.
When your window looks like the window above, click Next. This will bring you to the
Filtering and Interpolation window.
8.4 Step 4. Filtering and Interpolation
The Filtering and Interpolation window has several functions. These include identifying
the years of monitor data that you want to use, choosing the particular monitors in these
Abt Associates Inc.
216
June 2010
-------
Ozone Analysis: Quick Start Tutorial
data that you want in your analysis, and (when calculating spatial fields) specifying the
interpolation method. Use the default settings pictured in the screenshot below.
Choose Ozone Design Values. Choose the years of design values that you want to use.
The default is to use a 5 year period (3 design values) that is centered about the base
emissions year. The default in MATS assumes an emissions year of 2002. Therefore,
the design value would be based on data from 2000-2002 up through a design value
based on data from 2002-2004. (That is, the Start Year is 2000-2002 and the End Year
is 2002-2004.)
Valid Ozone Monitors. Identify "valid" monitors — that is, those monitors that you want
to include in the analysis. The defaults are that monitors should have at least one valid
design value period; and, are within 25 kilometers of a model grid cell. You can also
specify that a monitor must have a particular design value (e.g., 2000-2002) to be valid,
however the default is to require none in particular.
Default Interpolation Method. Choose the interpolation method — that is, the method to
combine the design values from different monitors into a single estimated design value.
This option is only used when generating estimates for a Spatial Field. Since we are only
generating Point Estimates, this set of options is not active.
Filtering/Interpolation
RRF/Spatial Gradient
Final Check
Choose Ozone Design Values
Start Year
2000-2002
E End Year
2002-200-1
3]
Valid Ozone Monitors
Minimum Number of design values 1
Max Distance from Domain [km] |25
Required Design Values
None selected
E
Default Interpolation Method
Inverse Distance Weights
I check to set a maximum interpolation distance [km]
3]
< Back
Next >
Cancel
When your window looks like the window above, click Next. This will bring you to the
RRF & Spatial Gradient window, where you can set parameters for the calculation of
Abt Associates Inc.
217
June 2010
-------
Ozone Analysis: Quick Start Tutorial
RRFs and spatial gradients.
8.5 Step 5. RRF & Spatial Gradient
The RRF and Spatial Gradient window has two sets of options.
The RRF Setup uses threshold values in the model data to identify the days to be used in
the calculation of relative response factors (RRFs). The details of this process are
somewhat involved and are described in detail in the next chapter. (See: the example
calculations.) A brief summary is the following: The default threshold is set to 85 ppb.*
If there are fewer than 10 model days at or above 85 ppb in the baseline scenario, then
MATS will lower the threshold in increments of 1 ppb, until there are at least 10 days at
or above this new, lower threshold. This process is continued, if needed, until a threshold
of 70 ppb is reached. By default, this is the lowest allowable threshold. If there are fewer
than 5 days at or above this threshold of 70 ppb, then the monitor site will be dropped.
The Spatial Gradient Setup identifies the model values that will be used in the
calculation of a Spatial Field. Since we are only generating Point Estimates, this set of
options is not active.
RRF and Spatial Gradient
r
RRF Setup:
Initial threshold value (ppb)
Minimum number of days in baseline at or above threshold
Minimum allowable threshold value (ppb)
Min number of days at or above minimum allowable threshold |~~
r Enable Backstop minimum threshold for spatial fields
Backstop minimum threshold for spatial fields
Subrange first day of ozone season used in RRF [T
Subrange last day of ozone season used in RRF
f Pair days based on high concentration instead of date.
Spatial Gradient Setup:
Start Value
End Value
1153
85
lo
70
~~!5
< Back
Next >
Cancel
IBS Desired output
I Data Input
¦ Filtering/Interpolation
When your window looks like the window above, click Next. This will take you to the
Final Check window, where you can verify the choices that you have made.
Abt Associates Inc.
218
June 2010
-------
Ozone Analysis: Quick Start Tutorial
* The default values in MATS are consistent with the recommended values in the EPA modeling
guidance (see section 14.1.1).
8.6 Step 6. Final Check
The Final Check window verifies the choices that you have made. For example, it makes
sure that the paths specified to each of the files used in your Configuration are valid.
Click on the Press here to verify selections button.
Desired output
Data Input
Filtering/Interpolation
RRF/Spatial Gradient
Final Check
Final Check
Verify inputs
Press here to verify your selections...
Checking...
Check OK. Press the finish button to continue..
Save Scenario
< Back Save Scenario & Run
Cancel
If you encounter any errors, go back to the choices you have previously made by clicking
on the appropriate part (e.g., Data Input) of the tree in the left panel, and then make any
changes required.
When your window looks like the window above, click either Save Scenario & Run or
Save Scenario. Save Scenario & Run will cause MATS to immediately run the scenario.
A temporary, new Running tab will appear (in addition to the Start, Map View and
Output Navigator tabs).
Abt Associates Inc.
219
June 2010
-------
Ozone Analysis: Quick Start Tutorial
4
MATS
EEf
tl
Utilities ' Help T
Start Map View Output Navigator | Running |
| Close ;|
Name | Last Message
Tutorial 03.asr Loading Ozone monitor data...0.156 s.
l| ***
When the calculations are complete, a small window indicating the results are Done will
appear. Click OK.
Done
OK
After clicking OK, the Output Navigator tab will be active. (The Running tab will no
longer be seen.) MATS will automatically load the output files associated with the .asr
configuration that just finished running.
Abt Associates Inc.
220
June 2010
-------
Ozone Analysis: Quick Start Tutorial
011®
Help T
Start Map View | Output Navigator I
Load
Extract All
Highlight file of interest and right-click to view options to Map, View, and Extract the data.
|Type
Name
Size
B Configuration/Log Files
|- Co nf i g u rati on Co nf i g u rati o n
Log File Run Log
!=}•Output Files
j- Tutorial 03-Ozone Monitors-monitor data temporally adjusted 2015 Monitor Network
Tutorial 03 - Ozone Monitors - county high monitoring sites, temporally adjusted 2015 Monitor Network
81 kb
1 kb
85kb
56kb
Stop Info
The next step (click here) shows you how to map your results with the Output Navigator.
For more details on mapping and other aspects of the Output Navigator, there is a
separate chapter on the Output Navigator.
8.7 Step 7. Load & Map Output
After generating your results, Output Navigator can be used to load and/or map them. If a
run just finished, the output files will already be loaded into output navigator.
If files from a previous run need to be loaded then click on the Load button and choose
the Tutorial 03.asr file.
Abt Associates Inc.
221
June 2010
-------
Ozone Analysis: Quick Start Tutorial
Open MATS Result file
My Recent
Documents
3
Desktop
My Documents
&
My Computer
%
My Network
Places
Look in: output
3
E &
Example Annual PM
|£|Exarinple Daily PM
ICj Example 03
Example Visibility
j Tutorial Annual PM
j Tutorial Daily PM
jTutorial 03
jTutorial Visibility
J^i] Example Annual PM.asr
±3 Example Daily PM.asr
Example 03,asr
Q Example Visibility.asr
j|i] Tutorial Annual PM.asr
^Tutorial Daily PM.asr
SI Tutorial 03.asr
] Tutorial Visibility,asr
File name:
Files of type:
Tutorial 03.as
MATS Result File
-
"3
Open
Cancel
Under Configuration/Log Files, you will see two files:
Configuration: keeps track of the assumptions that you have made in your analysis.
Log File : provides information on a variety of technical aspects regarding how a results
file (*.ASR) was created.
Under Output Files you will see:
Tutorial 03 - Ozone Monitors - monitor data, temporally adjusted 2015: contains
forecasted values and the monitor data used.
Tutorial 03 - Ozone Monitors - county high monitoring sites, temporally adjusted 2015:
contains forecasted values and the monitor data used for the monitor with the highest
levels in the county.
Right-click on the file Tutorial 03 - Ozone Monitors - monitor data temporally adjusted
2015. This gives you three options: Add to Map, View, and Extract. Choose the Add to
Map option.
Abt Associates Inc.
222
June 2010
-------
Ozone Analysis: Quick Start Tutorial
B mats
Help T
Start Map View | Output Navigator I
~11®
Load
Extract All
Highlight file of interest and right-click to view options to Map, View, and Extract the data.
A[Type
Name
Size
B Configuration/Log Files
j- Configuration
Log File
B Output Files
Configuration
Run Log
Tutorial 03 - Ozone Monitors - county high monitoring sites, temporally adjusted 2015 Monitor Network
Tutorial 03 - Ozone Monitors - monitor data
View
Extract
ted 2015
Monitor Network
81 kb
1 kb
85kb
Stop Info
This will bring up the Map View tab.
Abt Associates Inc.
223
June 2010
-------
Ozone Analysis: Quick Start Tutorial
Start | Map View | Output Navigator
j: Data Loaded ||
0
[3
Standard Layers »
Tutorial 03 - Ozone Monitors -
Long:-184.08737. Lot: 31.29677 ***
Extent: Min(-159.661,2.334) Max(-&"«M,51.197)
3 Monitors -
Maryland
New England
Southern California
Texas
Washington DC
MATS
EHIB
Long:-185.96632, Lat: 30.58705 **"
Extent: Min(-159.661,2.334) Max(-fr"*H51.197)
Start | Map View | Output Navigator
To view an enlarged map, use the Zoom to an area Task Bar button on the far left.
Choose the Continental US.
Full Extent
Standard Layers »
Edit Zoom Frames
Add Current View to List
Abt Associates Inc.
224
June 2010
-------
Ozone Analysis: Quick Start Tutorial
^ V Si 0
JiData Loaded i|
@ ° Tutorial 03-Ozone Monitors -
Standard Layers T
~ US States
V mats
BBS
Long:-126.35362, Lat: 50.44844 ***
Extent: Min(-122.607,21.608) Max(^.820,45.788)
Start | Map View | Output Navigator
To more easily view the location of monitors in particular states, uncheck US Counties
using the Standard Layers drop down menu on the far right of the Task Bar. Your
window should look like the following:
Help "
Zoom in further on the Eastern US using the Zoom in button on the Task Bar. This allows
you to view the results more closely. A dashed line surrounds the area that you have
chosen and should look something like the following:
Abt Associates Inc.
225
June 2010
-------
Ozone Analysis: Quick Start Tutorial
I Help »
Start Map View Output Navigator
Long:-74.94675, Lat: 25.04686 ***
Extent: Min(-122.607,21.608) Max(-***.820,45.788)
' —1 —IIstop lnf0
Right click on the " Tutorial 03 - Ozone Monitors -"layer in the panel on the left side of the
window. Choose the Plot Value option.
Start | Map View | Output Navigator
Stop Info
Long:-109.78569, Lat: 28.76591 ***
Extent: Min(-109.583,26.788) Max(-»*411,41.980)
+ T + K.' jr^S) ^ Standard Layers
| Data Loaded | .
@ ° Tutorial 03- Q*r>no Mnmtnre-
Remove
Export as CSV File
Abt Associates Inc.
226
June 2010
-------
Ozone Analysis: Quick Start Tutorial
This will bring up Shape Class Breaks window. In the Value drop-down list, choose the
variable "f o3 civ" — this is forecasted ozone design value for 2015.
Shape Class Breaks
Layer Name: Tutorial 03 - Ozone Monitors -- county high monitoring
Value:
Date
f o3 dv
2004
Bins C Unique Values
E
Class Count: 5
Start Color
Marker Sizing: [0~~ ^
End Color
Clear Breaks | V Apply X Close |
Click Apply and then click Close. This will bring you back to the Map View window.
Help
Start | Map View | Output Navigator
^ Standard Layers -
Data Loaded
@ ° Tutorial 03-Ozone Monitors -
@ • f_o3_dv-9 to 58.2
@ • f_o3_dv 58.3 to 68.6
@ ® f_o3_dv 68.6 to 72.3
@ • _f_o3_dv 7ZA to 71A
@ ° f 03 dv 77.5 to 96.3
Xry/j":
0 f
• Js f1• • f
"M yyj
-------
Ozone Analysis: Quick Start Tutorial
rrf relative response factor used to forecast the ozone design value;
ppb \ value of the threshold used;
day: number of days at or above the threshold.
This is just a brief summary of the mapping possibilities available. For more details, there
is a separate chapter on the Map View. The next step is to go to the Output Navigator to
view the data in a table format.
8.8 Step 8. View & Export Output
After mapping your results, click on the Output Navigator tab, so that you can then view
the data in a table. Right-click on the file Tutorial 03 - Ozone Monitors - monitor data
temporally adjusted 2015. This gives you three options: Add to Map, View, and Extract.
Choose the View option.
Help
Start Map View | Output Navigator I
Load
Extract All
Highlight file of interest and right-click to view options to Map, View, and Extract the data.
|Type
| Name
El Configuration/Log Files
I | Configuration Configuration
I | Log File Run Log
G Output Files
Tutorial 03 - Ozone Monitors — county high monitoring sites, temporally adjusted 2015 Monitor Network
Monitor Network
Size
81 kb
1 kb
56kb
]| Stop Info
This will bring up a Monitor Network Data tab. The upper left panel allows you to view
the ID and latitude and longitude of the monitors in your data — at the right of this panel
there is a scrollbar with which you can locate any particular monitor of interest. The lower
left panel allows you to view the other variables in the data.
Abt Associates Inc.
228
June 2010
-------
Ozone Analysis: Quick Start Tutorial
Help
Start Map View Output Navigator | Monitor Network Data |
Tutorial 03 - Ozone Monitors — monitor data, temporally adjusted 2015
Show All | or select a particular location to see c
I type
[long
010030010
010270001
010331002
010510001
010550011
010690004
010730023
ni n-701 nni
30.498001
33.281261
34.760556
32.4985667
33.904039
31.1906565
33.553056
AOCCCC
-87.8814123
-85.8021817
-87.650556
-86.1365871
-86.0538672
-85.423117
-86.815
ojLHL
Is
Select Quantities that must be >= 0
~ b_o3_dv
~ f_o3_dv
~ referencecell
J rrf
~ ppb
~ days
Export Export this data to CSV
id
date
b_o3_dv
f_o3_dv
referencece
rrf
ppb
days
i
010030010
2004
78.0
68.8
95023
0.8825
85.0
11.0
010270001
2004
79.3
62.7
108051
0.7909
71.0
11.0
010331002
2004
-7.00
-9.00
92063
0.7642
71.0
11.0
010510001
2004
76.7
63.2
106043
0.8251
70.0
9.00
f
010550011
2004
75.0
58.0
105056
0.7738
73.0
10.0
010690004
2004
-7.00
-9.00
113032
-9.00
70.0
1.00
nm73fin?3
?nn4
77 0
60?
rnnnR?
0 7828
81 n
11 n
Stop Info
The default option is to show all of the data in the lower left panel. If, however, you want
to just view the data for a particular monitor — in this example, monitor ID = "010331002"
— use the scrollbar (if needed) and then highlight this monitor. MATS will then display
the values for this monitor in the bottom panel.
Abt Associates Inc.
229
June 2010
-------
Ozone Analysis: Quick Start Tutorial
To view all of the data again, click on the Show All button.
Help
Start Map View Output Navigator | Monitor Network Data I
Tutorial 03 - Ozone Monitors — monitor data, temporally adjusted 2015
Close
Show All | or select a particular location to see data.
1 I type
010030010
30.4980011
[long
-87.8814123
010270001
010331002
010510001
010550011
010690004
010730023
m m-01 nno
33.2812611
34.760556
32.4985667 j
33.904039
31.1906565
33.553056
Aocncc
-85.8021817
-87.650556
-86.1365871
-86.0538672
-85.423117
-86.815
pi; m
IE
Select Quantities that must be >= 0
b o3_dv
H referencecell
rrf
ppb
days
Export Export this data to CSV
id
(date
b_o3_dv
f_o3_dv
referencece
rrf
ppb
days
n
010030010
12004
78.0
68.8
95023
0.8825
85.0
11.0
010270001
2004
79.3
62.7
108051
0.7909
71.0
11.0
010331002
12004
-7.00
-9.00
92063
0.7642
71.0
11.0
010510001
12004
76.7
63.2
106043
0.8251
70.0
9.00
010550011
2004
75.0
58.0
105056
0.7738
73.0
10.0
010690004
2004
-7.00
-9.00
113032
-9.00
70.0
1.00
F
010730073
l?0f]4
77 0
60?
10005?
0 7fi?R
81 0
11 0
J | Stop Info
To eliminate missing values (denoted by negative numbers in the lower panel), check one
Abt Associates Inc.
230
June 2010
-------
Ozone Analysis: Quick Start Tutorial
or more boxes in the panel in the upper right of the window. For example, to eliminate any
monitors that do not have a ozone design value forecast, check the forecasted ozone design
value variable "f o3 dv". MATS will automatically drop these values. (Note that the
monitor that we previously highlighted [monitor ID = "010331002] has now dropped out
of the display.)
Help '
Start M ap Vi e w Qutput Navigator | Monitor Network Data I
Close
id
J Data [
Refresh ~| Select location and press refresh to see data.... j; ShowAiH|
170230001
170310001
170310032
170310050
170310063^
170310064
170310072
1 nmc
H type
- lat
39.210883
41.672745
41.755833
41.709561
41.877222
41790833
41.895833]"
a-i nc/ii c-j
' [long
Select Quantities that must be >= 0
~ b_o3_dv
*0~
-87.668416
-87.732457 C
~ referencecell
~ rrf
~ ppb
O days
-87.634444
-87.607595 ¦
Q-; croci 1
Export Export this data to CSV
id
010030010
010270001
010331002
010510001
010550011
010730023
010731003
010731005
[^lldate r^1|b_o3_dv[^l|f_o3_dv [z |referenc([^l|rrf [^"llppb Fridays [^1|
2005
2005
2005
2005
2005
2005
2005
2005
77.7
77.7
71.0
75.0
73.0
75.7
78.0
79.2
68.6
_61.4
54.2
61.8
56.5
59.2
J52.4
60.5
95023
108051
92063
106043
105056
100052
99052
99050
0.883
0.791
0.764
0.825
0.774
0.783
0.801
0.764
85.0
71.0
71.0
70.0
73.0
81.0
81.0
80.0
11.0
11.0
11.0
9.00
10.0
11.0
11.0
10.0
Stop
JE
Info
Click the Export button and save the file as "03 Tutorial No Negative Forecasts." (It is
unnecessary to add an extension. MATS automatically saves the file as a CSV text file and
adds a ".csv" extension to your file name.) View the file in Excel.
Abt Associates Inc.
231
June 2010
-------
Ozone Analysis: Quick Start Tutorial
r*
E2 Microsoft Excel - (
)3 Tutorial No Negative Forecasts.csv
Bt
=pl
0
File Edit
View
Insert Format Tools Data Window
Help
-
I? X
a
» 10 -
w
b i
— — —
too +"°8 iF tp <3* - A .
r
»
~
~ & O ^
n
m a y * n
! >o - Z
A I Z
z\ A
i a ioo% -
A1
T
fx id
A
B
C
D
E
F
G
H
Tj
i
id II d at e
b o3 dv
f o3 dv referencecell
rrf
PPb
days
2
10030010
2004
78
68.8
95023
0.8825
85
11
3
10270001
2004
79.3
62.7
108051
0.7909
71
11
4
10510001
2004
76.7
63.2
106043
0.8251
70
9
5
10550011
2004
75
58
105056
0.7738
73
10
6
10730023
2004
77
60.2
100052
0.7828
81
11
7
10731003
2004
79
63.2
99052
0.8007
81
11
8
10731005
2004
80
61.1
99050
0.7644
80
10
9
10731009
2004
81.3
62.9
96051
0.7748
75
11
10
10731010
2004
72.5
55.7
102053
0.7687
77
11
11
10732006
2004
83.7
65.1
100051
0.7783
81
10
12
10735002
2004
78.7
62.2
101054
0.7904
77
10
13
10735003
2004
79.7
64
99055
0.8042
78
10
14
10736002
2004
78.7
62.1
100053
0.7892
80
10
jj r
M i
~ m \ 03 Tutorial No Negative Forecas/ |<
l
l^ead^^
NUM
For additional details on generating ozone results, see the chapter on Ozone Analysis:
Details. For additional details on viewing data, see the View Data section in chapter on
the Output Navigator
Abt Associates Inc.
232
June 2010
-------
Ozone Analysis: Details
9 Ozone Analysis: Details
MATS can forecast design values at ozone monitor locations — these forecasts are referred
to as Point Estimates. MATS can also use a variety of approaches to calculate design
values for a Spatial Field. A Spatial Field refers to a set of values comprising calculations
for each grid cell in a modeling domain from Eulerian grid models such as CMAQ and
CAMx.
The set of choices involved in calculating either Point Estimates or a Spatial Field can be
fairly involved, so MATS keeps track of these choices using a Configuration. When you
begin the process of generating ozone estimates, MATS provides an option to start a new
Configuration or to open an existing Configuration.
Configuration Management
(* iCreateNew Configurations
C Open Existing Configuration
Select your option and then click Go.
MATS will then step you through a series of windows with choices for your analysis.
Choose Desired Output. Choose whether you want Point Estimates, estimates for a
Spatial Field, or both.
Data Input. Specify the air quality modeling and ambient monitoring data that you want
to use. Specify which model grid cells will be used when calculating RRFs at monitor
locations.
Filtering Interpolation. Choose the years of monitoring data. Identify valid monitors.
Define the interpolation approach to be used (when calculating a Spatial Field).
RRF and Spatial Gradient. Specify the daily ozone values that will be used in the
calculation of RRFs and Spatial Gradients.
Final Check. Verify the selections that you have made.
9.1 Choose Desired Output
MATS lets you choose to generate Point Estimates, which refer to forecasts made at fixed
locations, such as monitors. MATS can also generate Spatial Fields, which refer to air
pollution estimates made at the center of each grid cell in a specified model domain. (For
example if the model domain has 20 columns and 30 rows, then there are 600 grid cells
for which MATS can generate estimates.) The Spatial Field estimates can be baseline
estimates or forecasts, generated with or without a gradient adjustment.
Go
Cancel
Abt Associates Inc.
233
June 2010
-------
Ozone Analysis: Details
By checking the box next to Automatically extract all selected output files, MATS will
create a separate folder with your chosen Scenario Name in the MATS "Output" folder,
and then export .CSV files with the results of your analysis. Alternatively, you can export
the results from the Output Navigator, but checking this box is a little easier.
Desired output
¦ Data Input
¦ Filtering/I interpolation
¦ RRF/Spatial Gradient
||9 Final Check
Choose Desired Output
Scenario Name: [Example 03
Point Estimates
Forecast
p T emporally-adjust ozone levels at monitors.
Spatial Field
Baseline
M Interpolate monitor data to spatial field
p Interpolate gradient-adjusted monitor data to spatial field.
Forecast
p Interpolate monitor data to spatial field. Temporally adjust ozone levels.
^Interpolate gradient-adjusted monitor data to spatial field. Temporally adjust.
Actions on run completion
W Automatically extract all selected output files
< Back
Next >
Cancel
9.1.1 Scenario Name
The Scenario Name allows you to uniquely identify each analysis that you conduct. It is
used in several ways.
Results file name. The results file is given the Scenario Name (e.g., Example 03.asr).
Note that the extension (ASR) is specifically designated just for MATS and can only be
used by MATS.
Organize output. In the Output folder, MATS will generate a folder using the Scenario
Name. MATS will use this folder as a default location for files generated with this
Scenario Name.
Abt Associates Inc.
234
June 2010
-------
Ozone Analysis: Details
i* C:\Program Files\Abt Associates\MATS\output |. |[~ |[X
Q C:\Program Files\Abt Associates\MATS\output
File Edit View Favorites Tools Help
@Back t ' J Search
Address
Folders
: b i£3 Documents and Settings
Q drvrtmp
fe ic3 ESP 1
! & My Downloads
Program Files
© £3 Abt Assoc
! a & Abt Associates
i Q MATS
(f^l configs
r O data
(£l help
Q maps
Folders
a go
Name
QExample 03
B] Example 03,asr
Size Type
File Folder
3,729 KB ASR File
Output file names. The output files generated will begin with the Scenario Name.
• MATS
Help -
raw
Start Map View | Output Navigator I
|; Load ~i| | Extract All | Highlightfile of interest and right-click to view options to Map. View, and Extractthe data.
Name
Type
Size
Configuration/Log Files
Configuration
Configuration
2kb
Log File
Run Log
1 kb
Output Files
Example 03 - Ozone Monitors - monitor data temporally adjusted 2015
Monitor Network
80kb
Example 03 - Ozone Monitors - county high monitoring sites, temporally adjusted 2015
Monitor Network
54kb
Example 03 - Spatial Field - interpolated monitor data temporally adjusted; gradient-adjusted monitor..
Monitor Network
3588kb
Stop
Info
Abt Associates Inc.
235
June 2010
-------
Ozone Analysis: Details
9.1.2 Point Estimates
The calculation of Point Estimates, or future-year ozone levels at monitors, has several
steps (the process is laid out in more detail in Sections 3 and 4 of the EPA modeling
guidance). The first step is to calculate the baseline value as a function of up to three
design values. The second step is to use model data to temporal adjust the baseline value.
Scenario Name : |Tutorial 03
Point Estimates
Forecast
T emporally-adjust ozone levels at monitors.
Spatial Field
Baseline
Interpolate monitor data to spatial field
P Interpolate gradient-adjusted monitor data to spatial field.
Forecast
P Interpolate monitor data to spatial field. Tennporally adjust ozone levels.
P Interpolate gradient-adjusted monitor data to spatial field. Temporally adjust.
Actions on run completion
I* ^Automatically extract all selected output file?
9.1.2.1 Baseline Ozone
The baseline ozone design value is the simple average of design values, where the average
carries one significant figure to the right of the decimal point. Generally, one should select
design value years that match the modeling data being used. The EPA modeling guidance
recommends using an average of the 3 design values periods which straddle the emissions
base year. For example, if the modeled emissions base year is 2002, then design values
from 2000-2002, 2001-2003, and 2002-2004 would be averaged. An average of design
values is, in effect, a weighted average of annual averages — 2002 is "weighted" three
times, 2001 and 2003 are weighted twice, and 2000 and 2004 are weighted once. This
creates a 5-year weighted average design value which is used to project future air quality
levels.
The default design value years in MATS are the periods 2000-2004. This assumes a model
base year of 2002. If the base year is not 2002, then the start and end design value period
should be adjusted.
Abt Associates Inc.
236
June 2010
-------
Ozone Analysis: Details
9.1.2.2 Temporally-Adjust Baseline Ozone
The first step in temporally adjusting baseline ozone involves identifying the model grid
cells near the monitor site. Next, MATS calculates the average of daily 8-hour average
maximum model values for both the baseline and future-year model runs, and then takes
the ratio of the two to calculate the RRF. Finally, MATS calculates the future-year design
value by multiplying the RRF with the baseline design value measured at the monitor.
The equation for temporally adjusting baseline ozone is as follows:
Monitor, flinire = Monitor) ¦ RRF.]
where:
Monitor; &ture = future-year ozone design value at monitor site i, measured in parts per
billion (ppb)
Monitor; = baseline ozone design value at monitor site i, measured in ppb
RRF; = relative response factor at monitor site i. The RRF is the ratio of the future 8-hour
daily maximum concentration predicted near a monitor (averaged over multiple days) to
the baseline 8-hour daily maximum concentration predicted near the monitor (averaged
over the same days).
NOTE: The calculation of the RRF involves a number of assumptions that are specified in
the RRF and Spatial Gradient window.
9.1.3 Spatial Field
A Spatial Field refers to air pollution estimates made at the center of each grid cell in a
specified model domain. For example, MATS might calculate ozone design values for
each grid cell in a modeling domain.
MATS calculates four types of ozone-related Spatial Fields:
Baseline - interpolate monitor data to spatial field. This is an interpolation of baseline
monitor values at each grid cell. MATS identifies the "neighbor" monitors for each grid
cell and then calculates an inverse-distance-weighted average of the monitor values at
each grid cell.
Baseline - interpolate gradient-adjusted monitor data to spatial field. This is an
interpolation of model-adjusted baseline monitor values at each grid cell. MATS
identifies the "neighbor" monitors for each grid cell, it adjusts the monitor values to
account for the modeled spatial gradient, and then calculates an inverse-distance-
weighted average of the monitor values.
Forecast - interpolate monitor data to spatial field. Temporally Adjust. This is an
interpolation of baseline monitor values at each grid cell that are then temporally
Abt Associates Inc.
237
June 2010
-------
Ozone Analysis: Details
adjusted to a future year. MATS calculates the Baseline - interpolate monitor data to
spatial field and multiplies it with a RRF.
Forecast - interpolate gradient-adjusted monitor data to spatial field. Temporally adjust.
This is an interpolation of model-adjusted baseline monitor values at each grid cell that
are then temporally adjusted to a future year. MATS calculates the Baseline -
interpolate gradient-adjusted monitor data to spatial field and multiplies it with a
RRF.
Details on the calculations are provided in the following sections.
Scenario Name : Example 03
Point Estimates
Forecast
T emporally-adjust ozone levels at monitors.
Spatial Field
Baseline
Interpolate monitor data to spatial field
Interpolate gradient-adjusted monitor data to spatial field.
Forecast
Interpolate monitor data to spatial field. Tennporally adjust ozone levels.
[Interpolate gradient-adjusted monitor data to spatial field. Temporally adjust.!
Actions on run completion
Automatically extract all selected output files
9.1.3.1 Baseline - interpolate monitor data to spatial field
Using modeling data for gradient scaling is fairly simple: MATS uses the model value for
the grid cell of interest and the model values for the grid cells containing the monitors to be
interpolated to the grid cell of interest. A general form of the equation is as follows:
n
GridcellEmbasaSjy!e = T Weight, ¦ Momtoi\ - Gradient Adjustment E
i-i
where:
Gridcell, baseline = baseline ozone concentration at unmonitored site E;
Abt Associates Inc.
238
June 2010
-------
Ozone Analysis: Details
Weight; = inverse distance weight for monitor i;
Monitor; = baseline ozone concentration at monitor i;
Gradient Adjustment; E = gradient adjustment from monitor i to unmonitored site E.
There are a variety of approaches that might be used to calculate the gradient adjustment.
As a default, MATS averages the five highest daily 8-hour values. The equation can then
be rewritten as follows:
" Moiie!£
Gridc€lIEMms = 2, Weight; ¦ Monitor, ¦ ——
,-l Modi-1,
where:
Model, baseline = baseline scenario, average of five highest daily 8-hour values at site E;
Model; baseline = baseline scenario, average of five highest daily 8-hour values at monitor
site i.
9.1.3.2 Baseline - interpolate gradient-adjusted monitor data to spatial field
Using modeling data for gradient scaling is fairly simple: MATS uses the model value for
the grid cell of interest and the model values for the grid cells containing the monitors to be
interpolated to the grid cell of interest. A general form of the equation is as follows:
n
Giidceth ias4lmt = Weight. ¦ Monitor - Giridient Adjitsniuvir. E
i-l
where:
Gridcell, baseline = baseline ozone concentration at unmonitored site E;
Weight; = inverse distance weight for monitor i;
Monitor; = baseline ozone concentration at monitor i;
Gradient Adjustment; E = gradient adjustment from monitor i to unmonitored site E.
There are a variety of approaches that might be used to calculate the gradient adjustment.
As a default, MATS averages the five highest daily 8-hour values. The equation can then
Abt Associates Inc.
239
June 2010
-------
Ozone Analysis: Details
be rewritten as follows:
_ . 7 77 " _ , _ ModelE ba:elim
GndcellE basetjBS = ^ Weightt ¦ Monitor;¦
i-l MOCteli,baitlm
-------
Ozone Analysis: Details
where:
Gridcell, baseline = baseline ozone concentration at unmonitored site E;
Weight; = inverse distance weight for monitor i;
Monitor; = baseline ozone concentration at monitor i.
NOTE: The RRF is calculated using the same approach as for Point Estimates (except
when a backstop threshold minimum is set rsee section 9.4.1.61Y
9.1.4 Ozone Output Variable Description
MATS generates up to three output files:
Ozone forecasts for all monitors. The name of this file is "Ozone Monitors — monitor
data, temporally adjusted yyyy.csv" with the Scenario Name appended at the beginning
and the forecast year is inserted at the end (e.g., "Example 03 — Ozone Monitors —
county high monitoring sites, temporally adjusted 2015.csv").
Ozone forecasts for the highest monitor in each county. The name of this file is "Ozone
Monitors — county high monitoring sites, temporally adjusted yyyy.csv" with the
Scenario Name appended at the beginning and the forecast year inserted at the end.
Spatial field forecasts. The name of this file is "Spatial Field — interpolated monitor
data, temporally adjusted; gradient-adjusted monitor data, temporally adjusted yyyy.csv"
with the Scenario Name appended at the beginning and the forecast year inserted at the
end.
The following sub-sections describe the variables in each file.
9.1.4.1 Ozone Monitors - monitor data, temporally adjusted 2015.csv
The table below describes the variables in the output file.
Variable Description
Jd The ID is a unique name for each monitor in a particular location. The default
value is the AIRS ID. (This is a character variable.)
_type Leave blank
lat Latitude in decimal degrees. Values in the northern hemisphere are positive, and
those in the southern hemisphere are negative.
long Longitude in decimal degrees. Values in the eastern hemisphere are positive, and
those in the western hemisphere (e.g., United States) are negative.
date The date represents the last year of the chosen design value periods (e.g., if a 5
year period is chosen, 2005 represents the 2001-2005 period ).
b_o3_DV Baseline design value
f_o3_DV Forecasted (future year) design value
Abt Associates Inc.
241
June 2010
-------
Ozone Analysis: Details
referencecell Identifier of the closest model grid cell centroid to the monitor.
rrf Relative response factor is the ratio of the future year modeled concentration
predicted near a monitor (averaged over multiple days) to the base year modeled
concentration predicted near the monitor (averaged over the same days).
ppb
days
_state_name
_county_name
Threshold value (measured in parts per billion) used in the rrf calculation
Number of days at or above the threshold value.
State name. (This is a character variable.)
County name. (This is a character variable.)
9.1.4.2 Ozone Monitors - county high monitoring sites, temporally adjusted
2015.csv
An example of this output file is as follows (with variable definitions in the table below):
Variable Description
Jd The ID is a unique name for each monitor in a particular location. The default value is
the AIRS ID. (This is a character variable.)
_type Leave blank
lat Latitude in decimal degrees. Values in the northern hemisphere are positive, and those
in the southern hemisphere are negative.
long Longitude in decimal degrees. Values in the eastern hemisphere are positive, and
those in the western hemisphere (e.g., United States) are negative.
date The date represents the last year of the selected design value periods (e.g., if a 5 year
period is selected, 2005 represents the 2001-2005 period ).
b_o3_DV Baseline design value
f_o3_DV Forecasted (future year) design value
referencecel Identifier of the closest model grid cell centroid to the monitor.
I
rrf Relative response factor is the ratio of the future year modeled concentration predicted
near a monitor (averaged over multiple days) to the base year modeled concentration
predicted near the monitor (averaged over the same days).
ppb Threshold value (measured in parts per billion) used in the rrf calculation
days Number of days at or above the threshold value.
_state_nam State name. (This is a character variable.)
e
_county_na County name. (This is a character variable.)
me
9.1.4.3 Spatial Field -- interpolated monitor data, temporally adjusted;
gradient-adjusted monitor data, temporally adjusted 2015.csv
An example of this output file is as follows (with variable definitions in the table below):
Variable Description
Abt Associates Inc.
242
June 2010
-------
Ozone Analysis: Details
Jd The ID is a unique name for each monitor in a particular location. The default value is
the column identifier multiplied by 1000 plus the row. (This is a character variable.)
_type Leave blank
lat Latitude in decimal degrees of the center of each grid cell. Values in the northern
hemisphere are positive, and those in the southern hemisphere are negative.
long Longitude in decimal degrees of the center of each grid cell. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g., United States)
are negative.
date The date represents the last year of the selected design value periods (e.g., if a 5 year
period is selected, 2005 represents the 2001-2005 period ).
ga_conc Modeled concentration (ppb) used for gradient adjustment (average of "start value"
and "end value")
i_b_o3 Interpolated (to spatial field) baseline concentration (ppb).
i_f_o3 Interpolated (to spatial field) future year concentration (ppb).
i_b_ga_o3 Interpolated (to spatial field) gradient adjusted baseline concentration (ppb).
i_f_ga_o3 Interpolated (to spatial field) gradient adjusted future year concentration (ppb).
ppb Threshold value (measured in parts per billion) used in the rrf calculation
days Number of days at or above the threshold value.
referencecellldentifier of the grid cell. (In the case of spatial fields, this is identical to the JD
variable.)
rrf Relative response factor is the ratio of the future year modeled concentration predicted
near a monitor (averaged over multiple days) to the base year modeled concentration
predicted near the monitor (averaged over the same days).
9.2 Data Input
In the Data Input window, you need to specify the air quality modeling and ambient
monitoring data that you want to use. In addition, you need to specify which model grid
cells will be used when calculating RRFs at monitor locations.
Abt Associates Inc.
243
June 2010
-------
Ozone Analysis: Details
B
Data Input
Monitor Data
~zone Data |mpleData\0ZQNE_MATS_input_99Q7-v2.csv ¦ ¦¦ |
Model Data
Baseline File fs~\SannpleData\ozone_mQdel_data_2QQ1.csv • |
Forecast File [s\S annpleD ata\ozone_rinQdel_data_2Q15.csv ¦"!
Using Model Data
Ternporal adjustment at monitor | 3x3 ^ | [Maximum
< Back
Next >
Cancel
9.2.1 Monitor Data
Monitor data should be in the form of a simple text file. The first row specifies the
frequency of the data (e.g., day). The second row presents comma-separated variable
names. The third row begins the data values. Below is an example of the monitor data file
format and descriptions of the variables in the file.
Format of Ozone Monitor Data
|DesignValue
_ID,_TYPE,LAT,
"010030010", ,
"010030010", ,
"010030010", ,
"010030010", ,
"010030010", ,
"010030010", ,
"010030010", ,
"010270001", ,
LONG,POC,DVYEAR,03,_S TATE_NAME,
30.497778,-87. 881389, 1, 1999,-9,
30. 497778,-87 .881389, 1,2000,-9,
30.497778,-87.881389,1,2001,-9,
30.497778,-87.881389,1,2002,82,
30.497778,-87.881389,1,2003,76,
30 . 497778,-87 .881389, 1,2004,76,
30 . 497778,-87 .881389, 1,2005,77,
33 .281111,-85 .802222, 1, 1999,88,
_COUNTY_NAME
"Alabama","Baldwin"
"Alabama","Baldwin"
"Alabama","Baldwin"
"Alabama","Baldwin"
"Alabama","Baldwin"
"Alabama","Baldwin"
"Alabama","Baldwin"
"Alabama","Clay"
Ozone Monitor Data Variable Descriptions
Variable Description
Abt Associates Inc.
244
June 2010
-------
Ozone Analysis: Details
_ID The ID is a unique name for each monitor in a particular location. The default
value is the AIRS ID. (This is a character variable.)
_TYPE Leave this blank.
LAT Latitude in decimal degrees. Values in the northern hemisphere are positive,
and those in the southern hemisphere are negative.
LONG Longitude in decimal degrees. Values in the eastern hemisphere are positive,
and those in the western hemisphere (e.g., United States) are negative.
DATE The time period of the monitor observation. As a convention, the date
represents the last year of the three-year design value period (e.g., 2001
represents the 1999-2001 design value).
03 Observed monitor value. Note that missing values are represented by a minus
nine (-9).
_STATE_NAME State name. (This is a character variable.)
_COUNTY_NAME County name. (This is a character variable.)
Note:
Character variables have names that begin with an underscore (i.e., and the
character values used can be kept with or without quotes. (If a character variable has an
embedded space, such as might occur with the name of a location, then use quotes.)
9.2.2 Model Data
The model data should be in the form of a simple text file. The first row specifies the
frequency of the data (e.g., day). The second row presents comma-separated variable
names. The third row begins the data values. The ozone model data should be the daily 8-
hour average maximum concentration in each grid cell. Below is an example of the model
data file format and descriptions of the variables in the file.
Format of Ozone Model Data
Day
_ID,_TYPE, LAT , LONG , DATE , 03
1001, "",28 .471949, -99 . 489582,20150501, 45 .2324
10 01, "",28 .4719 49, -99 . 489582,20150502, 42 . 6581
10 01,"",28.4719 49, -99.489582,20150503, 47 . 4534
10 01,"",28 .4719 49, -99 . 489582,20150504,51. 9678
10 01,"",28 .4719 49, -99 . 489582,20150505,53 . 6575
10 01, "",28 .4719 49, -99.489582,20150506, 47 .1936
10 01, "",28 .4719 49, -99 . 489582,20150507, 48 .3454
10 01,"",28 .4719 49, -99 . 489582,20150508, 49 .5464
10 01,"",28 .4719 49, -99.489582,20150509,34.3454|
Ozone Model Data Variable Descriptions
Variable Description
Abt Associates Inc.
245
June 2010
-------
Ozone Analysis: Details
ID
The ID is a unique number for each model grid cell in the air quality model domain. It
is generally based on the column and row identifiers from the air quality modeling
domain. The default convention is to calculate the ID by multiplying the column
identifier by one thousand (1000) and adding the row identifier. (This is a character
variable.)
_TYPE
LAT
Leave this blank.
Latitude in decimal degrees of the center of each grid cell. Values in the northern
hemisphere are positive, and those in the southern hemisphere are negative.
LONG
Longitude in decimal degrees of the center of each grid cell. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g., United States)
are negative.
DATE
03
The time of the monitor observation. The day is represented in the yyyymmdd format
Modeled ozone concentration (8-hour average daily maximum).
Note:
Character variables have names that begin with an underscore {i.e., "_"), and the
character values used can be kept with or without quotes. (If a character variable has an
embedded space, such as might occur with the name of a location, then use quotes.)
9.2.2.1 EPA Default Model Data
The example model output dataset in MATS comprises daily 8-hour average maximums
from May-September at 12km resolution. The baseline year for the modeling is 2001 and
the future year is 2015.
The RRF for a monitor is calculated from "nearby" model grid cells. For purposes of this
calculation, a monitor is assumed to be at the center of the cell in which it is located, and
this cell is at the center of an array of "nearby" cells.
The number of cells considered "nearby" a monitor is a function of the size of the grid
cells used in the modeling. In the example case of a 12 km grid, EPA uses as a default 3x3
array of grid cells (see section 3.2 of the EPA modeling guidance for more details).
9.2.3 Using Model Data
Abt Associates Inc.
June 2010
246
-------
Ozone Analysis: Details
H5| Desired output
Data Input
¦ Filtering/Interpolation
¦ RRF/Spatial Gradient
ESI Final Check
0
Data Input
Monitor Data
Ozone Data
SampleData\OZONE_ASIP_input_97-Q5.csv ¦¦¦"]
Model Data
Baseline File
Forecast File
Using Model Data
Temporal adjustment at monitor
\SampleData\ozone_model_data_2001 .csv 3D
\SampleData\ozone_model_data_2015.csv 3D
1x1
5x5
7x7
3x3 -
[Maximum 3]
< Back
Next >
Cancel
With the array size determined, MATS gives you two options for how you might use the
modeling data.
Maximum. For each day of modeling data, MATS will identify the highest 8-hour daily
maximum among the grid cells in the chosen array. In the case of a 3x3 array, MATS
will identify the highest daily 8-hour average maximum from among the nine "nearby"
grid cells for each day and for each monitor site.
Mean. For each day of modeling data, MATS will average the 8-hour daily values for
the grid cells in the chosen array. In the case of a 3x3 array, MATS will average nine
values.
Abt Associates Inc.
247
June 2010
-------
Ozone Analysis: Details
0
¦ Desired output
Data Input
¦¦¦iltering/lnterpolation
Data Input
« RRF/Spatial Gradient
¦ Final Check
Monitor Data
Ozone Data SampleData\OZONE_ASIP_input_97-Cl5.csv ¦¦¦ |
Model Data
Baseline File \SampleData\ozone_model_data_2001 .csv ¦¦¦ |
Forecast File \SampleData\ozone_model_data_2015.csv ¦¦¦ |
Using Model Data
Temporal adjustment at monitor I 3x3 T | [Maximum T
Mean
1
Maximum
< Back
Next >
Cancel
The default choice for the ozone analysis in MATS is to use the maximum value among
the array of grid cells, when calculating temporally-adjusted ozone levels at each monitor.
NOTE: For monitors on the border of a modeling domain — where it may not be possible
to have a full set of neighbors — MATS uses the available modeling data.
9.2.3.1 Nearby Monitor Calculation - Example 1
Given:
(1) Four primary days have been simulated using baseline and future emissions.
(2) The horizontal dimensions for each surface grid cell are 12 km x 12 km.
(3) In each of the 9 grid cells "near" a monitor site I, the maximum daily predicted future
concentrations are 87.2, 82.4, 77.5, and 81.1 ppb.
(4) In each of the 9 grid cells "near" a monitor site I, the maximum daily predicted baseline
8-hour daily maximum ozone concentrations are 98.3, 100.2, 91.6, and 90.7 ppb.
Find:
The site-specific relative response factor for monitoring site I, (RRF)I
Solution:
Abt Associates Inc.
248
June 2010
-------
Ozone Analysis: Details
(1) For each day and for both baseline and future emissions, identify the 8-hour daily
maximum concentration predicted near the monitor. Since the grid cells are 12 km, a 3 x 3
array of cells is considered "nearby."
(2) Compute the mean 8-hour daily maximum concentration for (a) future and (b) baseline
emissions. Using the information from above, (a) (Mean 8-hr daily max.)future = (87.2 +
82.4 + 77.5 + 81.1 )/4 = 82.1 ppb and (b) (Mean 8-hr daily max.)baseline = (98.3 + 100.2 +
91.6+ 90.7)/4 = 95.2 ppb
(3) The relative response factor for site I is
(RRF)I = (mean 8-hr daily max.)future/(mean 8-hr daily max.)baseline = 82.1/95.2 =
0.862.
Figure 3.1. Choosing Ozone Predictions To Estimate RRF's
(a) Predictions With Baseline Emissions
Day 1 Day 2 Day 3 Day 4
97.2
95.5
96.2
100.2
98.5
98.1
87.8
90.1
89.9
85.9
87.9
88.9
97.1
95.2
89.1
100.0
99.1
97.3
90.9
91.6
88.7
87.9
90.5
90.7
97.2
98.3
97.6
99.5
95.4
97.9
88.5
89.4
90.2
86.9
87.3
88.4
98.3 100.2 91.6 90.7
Mean Baseline Ozone Concentration = (98.3 + 100.2 + 91.6 + 90.7) / 4 = 95.2 ppb
Day 1
86.1
85,4
86.1
86.2
84.5
84.;
85.8
87.2
86,S
(b) Predictions With Future Emissions
Day 2 Day 3
Day 4
82.2
80.8
81.2
72.1
76.1
75.5
75.4
78.8
79.8
82.4
79.9
80.5
74.6
77.5
74.3
80.8
79.5
80.9
81.4
77.8
80.1
76.9
77.4
75.6
80,4
76.9
81.1
87,2 82.4 77.5 81.1
Mean Future Ozone Concentration = (87.2 + 82.4 + 77.5 + 81.1) I 4 = 82.1 ppb
9.3 Filtering and Interpolation
The Filtering and Interpolation window allows you to choose the years of monitoring
Abt Associates Inc.
249
June 2010
-------
Ozone Analysis: Details
data that you will use in your analysis. MATS allows you to specify the rules to determine
the monitors that you will use. And in the case of calculating Spatial Fields, it allows you
to define the interpolation method that MATS will use.
I Desired output
¦ Data Input
Filtering/Interpolation
¦ RRF/Spatial Gradient
Filtering and Interpolation
Choose Ozone Design Values
Start Year
j t End Year 12002-2004 -
Valid Ozone Monitors
Minimum Number of design values |1
Max Distance from Domain [km] |25
Required Design Values [None selected
Default Interpolation Method
I Inverse Distance Weights
f" check to set a maximum interpolation distance [km]
"3
103
< Back
Next >
Cancel
9.3.1 Choose Ozone Design Values
Choosing the Start Year and the End Year defines the range of the ozone design values
that will be used in the calculation of the baseline ozone level. You can vary the number
of design values used in this calculation.
The database that comes with MATS has design values periods from (1997-1999)-(2005-
2007). The default approach in MATS is to average 3 design value periods. For example,
if the modeling base year is 2002, then you would use the design values from 2000-2002,
2001-2003, and 2002-2004. The Start Year is set to 2000-2002 and the End Year is set
to 2002-2004.
Abt Associates Inc.
250
June 2010
-------
Ozone Analysis: Details
Choose Ozone Design Values
Start Year
2000-2002 -
EndYear |2002-2004
3
1097-1999
Valid Ozone Monitors
1998-2000
1999-2001
Minimum Number of design values
2000-2002
2001-2003
2002-2004
2003-2005
Max Distance from Domain [km]
Required Design Values
None selected
Default Interpolation Method
"3
Inverse Distance Weights
r check to set a maximum interpolation distance [km]
100
9.3.2 Valid Ozone Monitors
MATS provides three choices for identifying monitors that are "valid" and thus included in
your analysis.
Minimum Number of Design Values. Specifies the minimum number of design value
periods that need to be included in the calculation of the baseline ozone design value (1,
2, or 3).
Max Distance from Domain [km]. Specifies how far a monitor may be from a model
grid cell and still be included in calculations that use that grid cell's model values. (This
is relevant for the calculation of RRFs and gradient-adjustments.)
Required Design Values. Specifies whether a particular design value period needs to
be valid for the calculations to be performed at that monitor.
Abt Associates Inc.
251
June 2010
-------
Ozone Analysis: Details
Choose Ozone Design Values
Start Year [2000-2002 End Year 12002-2004
Valid Ozone Monitors
Minimum Number of design values 1
Max: Distance from Domain [km] |25
Required Design Values |None selected
Default Interpolation Method
Inverse Distance Weights 3
r check to set a maximum interpolation distance [km] |100
9.3.2.1 Minimum Number Design Values
The Minimum Number of Design Values specifies the minimum number of design
values that need to be available in the potential range of design values specified by the
Start Year and End Year. Monitors that do not meet the minimum are excluded from the
calculation of baseline ozone levels.
Recall that the baseline ozone level is an average of one or more design values. The
number of design values available for this calculation will typically be either 1, 2, or 3
design value periods. The default option is to require that one design value be available in
the specified range.
Abt Associates Inc.
252
June 2010
-------
Ozone Analysis: Details
Choose Ozone Design Values
Start Year [2000-2002 End Year 12002-2004
Valid Ozone Monitors
Minimum Number of design values j]
Max: Distance from Domain [km] |25
Required Design Values |None selected
Default Interpolation Method
Inverse Distance Weights 3
r check to set a maximum interpolation distance [km] |100
Example 1: Point Estimates
When calculating ozone levels at monitors, if MATS finds that a monitor has an
insufficient number of valid design values (e.g., one required but none available in the
specified range), then MATS will set the baseline level and forecast values to missing (
denoted by a negative number).
Example 2: Spatial Field
The calculation of the ozone level in each grid cell of a Spatial Field involves multiple
monitors. MATS uses Voronoi Neighbor Averaging to identify "neighboring" monitors
from the available set of valid monitors, and then calculates an inverse-distance weighted
average of the baseline ozone levels from these neighbors. The Minimum Number of
Design Values determines which monitors are "valid" — that is, those monitors that will be
included in the calculation.
9.3.2.2 Max Distance from Domain
The option to specify the Max Distance from Domain allows you to choose how far (in
kilometers) a monitor may be from the center of a model grid cell and still be included in
calculations that use that grid cell's model values.
Abt Associates Inc.
253
June 2010
-------
Ozone Analysis: Details
Choose Ozone Design Values
Start Year [2000-2002 End Year 12002-2004
Valid Ozone Monitors
Minimum Number of design values |1
Max: Distance from Domain [km]
Required Design Values |None selected
Default Interpolation Method
Inverse Distance Weights 3
r check to set a maximum interpolation distance [km] |100
Example 1: Point Estimates
When calculating ozone levels at monitors, if MATS finds that a monitor is further than
the specified Max Distance from Domain, then MATS will drop the monitor from the
analysis. For example, if the Max Distance from Domain is set to 25 kilometers and the
model domain includes the continental U.S., then monitors in Alaska, or other locations far
from the model domain are excluded from calculations involving model data. If an
extremely large distance is specified, say 10,000 kilometers, then all monitors would be
included, regardless of the model domain location
Example 2: Spatial Field
The calculation of the ozone level in each grid cell of a Spatial Field can involve multiple
monitors. MATS uses Voronoi Neighbor Averaging to identify "neighboring" monitors
from the available set of valid monitors, and then calculates an inverse-distance weighted
average of the baseline ozone levels from these neighbors. The Max Distance from
Domain determines which monitors are "valid" — that is, those monitors that will be
included in the calculation.
9.3.2.3 Required Design Values
Using the Required Design Value drop-down list, you can specify that a particular design
value must be available at each monitor included in the analysis. If you want to use all
Abt Associates Inc. June 2010
254
-------
Ozone Analysis: Details
monitors that have a valid design value for 2001-2003, then MATS will only include those
monitors that have this valid design value. The default is to choose None Selected.
Choose Ozone Design Values
Start Year 12000-2002 j^] End Year |2002-2004
Valid Ozone Monitors
Minimum Number of design values [T
Max Distance from Domain [km] 25
Required Design Values
None selected
~
r 1 997-1999 1
Default Interpolation Method
r1998-2000
r1999-2001
r 2000-2002
[inverse Distance Weights
| check to set a maximum interpol
r 2001-2003
r 2002-2004
r 2003-2005
9.3.3 Default Interpolation Method
The Default Interpolation Method panel allows you to choose how you will interpolate.
or combine, the values from different monitors. One approach is to use Inverse Distance
Weights. This means that the weight given to any particular monitor is inversely
proportional to its distance from the point of interest. A second approach is Inverse
Distance Squared Weights, which means that the weights are inversely proportional to the
square of the distance. And the third approach is Equal Weighting of Monitors. The
default approach for ozone is Inverse Distance Weights.
Abt Associates Inc.
255
June 2010
-------
Ozone Analysis: Details
Choose Ozone Design Values
Start Year 2000-2002
"3 End Year |2002-2004
Valid Ozone Monitors
Minimum Number of design values 11
Max Distance from Domain [km] |25
Required Design Values
None selected
Default Interpolation Method
| Inverse Distance Weights
Equal Weighting of Monitors
Inverse Distance Weights
Inverse Distance Squared Weights
When interpolating monitor values, MATS allows you to identify the monitors you want to
use based on their distance away from the point of interest (e.g., the center of a grid cell).
The first step in the interpolation process is to identify the monitors that are nearby, or
neighbors, for each point of interest. The next step is to determine the distance from the
nearby monitors to the point of interest.
The default approach is to include all valid monitors (i.e., those that satisfy the three
criteria in the Valid Ozone Monitors panel), regardless of distance. If you want to limit the
use of monitors based on distance, check the box next to check to set a maximum
interpolation distance, and then specify a distance (in kilometers). A distance of one
hundred (100) kilometers means that any monitors further than 100 kilometers can no
longer be used in the interpolation. If a point of interest has no monitors within the
specified distance, then no value is calculated. The default is to leave this box unchecked.
Abt Associates Inc.
256
June 2010
-------
Ozone Analysis: Details
Choose Ozone Design Values
Start Year 2000-2002 End Year |2002-2004
Valid Ozone Monitors
Minimum Number of design values 1
M ax Di stan ce fro rn D ~ rn ai n [km] 12 5
Required Design Values None selected Z
Default Interpolation Method
Inverse Distance Weights 3
F icheckto set a maximum interpolation distance [km]!
100
4i
M-
9.4 RRF and Spatial Gradient
In calculating an ozone RRF or a Spatial Gradient, typically, not all of the model data are
used. In the case of RRFs, daily values falling below specified thresholds can be excluded
from the calculation (e.g., RRF Calculation - Example 1). In the case of a spatial gradient,
MATS be be setup to follow the same thresholds as used for point estimates or if a valid
result is needed in all grid cells, a Backstop minimum threshold can be used (e.g., RRF
Calculation Spatial Gradient with Backstop Threshold - Example 6). You can also specify
a sub-range of days included in the model data files and base and future year model days
can be paired by high concentration instead of by date. MATS also averages a user-
specified range of values to calculate gradient adjustments (e.g., Spatial Gradient
Calculation - Example 1).
Abt Associates Inc.
257
June 2010
-------
Ozone Analysis: Details
Desired output
RRF and Spatial Gradient
RRF Setup:
Initial threshold value (ppb)
Minimum number of days in baseline at or above threshold
Minimum allowable threshold value (ppb)
Min number of days at or above minimum allowable threshold [
|~~ Enable Backstop minimum threshold for spatial fields
Backstop minimum threshold for spatial fields
Subrange first day of ozone season used in RRF
Subrange last day of ozone season used in RRF
f Pair days based on high concentration instead of date.
Spatial Gradient Setup:
Start Value
End Value
r~
1153
85
lo
~70
~~!5
60
"3
< Back
Next >
Cancel
H Nltering/lnterpolation
RRF/Spatial Gradient
B9 Final Check
9.4.1 RRF Setup
The RRF Setup involves four variables that specify the thresholds and the numbers of days
above the thresholds — Initial threshold value; Minimum number of days in baseline at
or above threshold; Minimum allowable threshold value; and Min number of days at
or above minimum allowable threshold.
The first step in calculating the RRF is to determines the number of days at or above the
Initial threshold value. If the number of days is above the Minimum number of days in
baseline at or above threshold, then MATS averages the 8-hour values for those grid
cells with at least this number. For example, MATS performs the following steps:
In the case of a 3x3 array, MATS identifies the highest daily 8-hour average maximum
from among the nine "nearby" grid cells for each day and for each monitor site. In the
case where there are 90 days of model outputs, MATS generates 90 daily values. NOTE:
MATS does this calculation separately for the baseline and future-year scenarios. As a
result two different grid cells in the baseline and future-year might be used to represent a
given day.
The default Initial threshold value is set to 85 ppb. The default Minimum number of
days in baseline at or above threshold is set to 10. If there are fewer than 10 days at or
above 85 ppb in the baseline scenario, then MATS lowers the threshold in increments of
1 ppb, until there are at least 10 days at or above this new, lower threshold. This process
is continued, if needed, until the Minimum allowable threshold value is reached. The
Abt Associates Inc.
258
June 2010
-------
Ozone Analysis: Details
default Minimum allowable threshold value is 70 ppb. MATS calculates the number
of days at or above the Minimum allowable threshold value. If there are fewer than
the Min number of days at or above minimum allowable threshold, then the monitor
site will be dropped. The default Min number of days at or above minimum
allowable threshold is 5.
Using the threshold established with the baseline scenario, MATS checks the daily 8-
hour maxima calculated for the baseline scenario, and sets to missing any daily value
falling below the threshold. For any day set to missing in the baseline scenario, MATS
also sets the corresponding day in the future-year scenario to missing.
For each monitor site, MATS averages the non-missing daily values for the baseline and
future-year scenarios, and then calculates the RRF as the ratio of the future-year average
to the baseline average.
You can also set a Backstop minimum threshold for spatial fields. As noted in Example
6 (below! the backstop minimum threshold allows the minimum threshold to be lowered
to a value below the Minimum allowable threshold value until the minimum number of
days is reached. The backstop threshold is only used for grid cells which do not have
enough days to meet the minimum number of days value with the minimum allowable
threshold. The backstop threshold does not change the calculation for grid cells that
already meet the minimum number of days.
Another option is to specify a Subrange of days to use in the model RRF calculation. The
start day and end day of the subrange are specified based on a count of the number of days
from the first day in the file. For example, if the model files contained data for June 1st-
August 30th, a subrange start and end day of "31" and "61" respectively would specify the
July lst-July 31st period.
There is also an option to pair days based on high concentration instead of date. This
may be useful for model runs where the future year meteorology is different than the base
year (such as climate modeling). MATS will pair the future year highest concentrations
with the highest concentrations in the base year file, regardless of date. The number of
days in the RRF will continue to be based on the threshold variables. For example, if 11
days are greater than the selected threshold in the base year, then the RRF will be
calculated based on the 11 highest base and future year concentration days.
9.4.1.1 RRF Calculation - Example 1
Assume the following default values:
Abt Associates Inc.
259
June 2010
-------
Ozone Analysis: Details
RRF Setup:
Initial threshold value (ppb)
85
Minimum number of days in baseline at or above threshold
10
Minimum allowable threshold value (ppb)
70
Min number of days at or above minimum allowable threshold |~~
5
Enable Backstop minimum threshold for spatial fields
Backstop minimum threshold for spatial fields
Subrange first day of ozone season used in RRF |1
4
Subrange last day of ozone season used in RRF pj 53
4
I- Pair days based on high concentration instead of date.
Spatial Gradient Setup:
Start Value
1
End Value
5
Assume that there are 15 days of data:
Baseline day Baseline Future day Future value
value
1
103
1
95
2
112
2
97
3
98
3
94
4
97
4
95
5
95
5
94
6
95
6
93
7
94
7
89
8
92
8
86
9
90
9
80
10
85
10
78
11
89
11
80
12
88
12
81
13
85
13
76
14
78
14
75
15
78
15
74
MATS will sort the values from high to low based on the Baseline values:
Abt Associates Inc.
260
June 2010
-------
Ozone Analysis: Details
Baseline day Baseline
value
2
112
1
103
3
98
4
97
5
95
6
95
7
94
8
92
9
90
11
89
12
88
10
85
13
85
14
78
15
78
Future day Future value
2
97
1
95
3
94
4
95
5
94
6
93
7
89
8
86
9
80
11
80
12
81
10
78
13
76
14
75
15
74
Note that Day 2 has the highest Baseline value. And note that the Future values are not
sorted high to low, and instead the Future days match the Baseline days.
When you compare these sorted data with the Initial threshold value of 85 ppb, note that
there are thirteen (13) Baseline values at or above this threshold. Since there are more
days than the ten (10) specified as the Minimum number of days in baseline at or above
threshold, MATS will use all 13 days.
MATS will take the top 13 days (highlighted in yellow) and then calculate separate
averages for the Control and Baseline values:
Control average = 87.5385 ppb
Baseline average = 94.0769 ppb.
The RRF equals the ratio of the Control to the Baseline:
RRF = 87.5385 / 94.0769 = 0.930
Note that we report the RRF with three digits after the decimal point. The calculation of
the Baseline and Control averages does not involve any rounding or truncation.
9.4.1.2 RRF Calculation - Example 2
Assume the following default values:
Abt Associates Inc.
261
June 2010
-------
Ozone Analysis: Details
RRF Setup:
Initial threshold value (ppb)
85
Minimum number of days in baseline at or above threshold
10
Minimum allowable threshold value (ppb)
70
Min number of days at or above minimum allowable threshold |~~
5
Enable Backstop minimum threshold for spatial fields
Backstop minimum threshold for spatial fields
Subrange first day of ozone season used in RRF |1
4
Subrange last day of ozone season used in RRF pj 53
4
I- Pair days based on high concentration instead of date.
Spatial Gradient Setup:
Start Value
1
End Value
5
Assume that there are 15 days of data:
Baseline day Baseline Future day Future value
value
1
100
1
95
2
100
2
97
3
98
3
94
4
97
4
95
5
95
5
94
6
95
6
93
7
90
7
89
8
85
8
86
9
84
9
80
10
83
10
78
11
83
11
80
12
83
12
81
13
79
13
76
14
78
14
75
15
78
15
74
MATS will sort the data from high to low based on the Baseline values:
Abt Associates Inc.
262
June 2010
-------
Ozone Analysis: Details
Baseline day Baseline
value
1
100
2
100
3
98
4
97
5
95
6
95
7
90
8
85
9
84
10
83
11
83
12
83
13
79
14
78
15
78
Future day Future value
1
95
2
97
3
94
4
95
5
94
6
93
7
89
8
86
9
80
10
78
11
80
12
81
13
76
14
75
15
74
Note that the Baseline values happen to stay in the same order. And note that the Future
values are not sorted high to low, and instead the Future days match the Baseline days.
When you compare these sorted data with the Initial threshold value of 85 ppb, note that
there are only eight (8) Baseline values (highlighted in yellow) at or above this threshold.
Since there are fewer days than the ten (10) specified as the Minimum number of days in
baseline at or above threshold, MATS will then lower the threshold by one ppb to 84 ppb.
There are nine (9) Baseline values at or above this lower threshold — still less than the
value of ten (10) that specified as the Minimum number of days in baseline at or above
threshold. MATS will lower the threshold again by one ppb to 83 ppb. At this point, there
are twelve (12) days at or above this threshold.
MATS will take the top 12 days:
Baseline day Baseline Future day Future value
value
1
100
1
95
2
100
2
97
3
98
3
94
4
97
4
95
5
95
5
94
6
95
6
93
7
90
7
89
8
85
8
86
9
84
9
80
10
83
10
78
Abt Associates Inc.
263
June 2010
-------
Ozone Analysis: Details
11
83
11
80
12
83
12
81
13
79
13
76
14
78
14
75
15
78
15
74
and then calculate separate averages for the Control and Baseline values:
Control average = 88.5000 ppb
Baseline average = 91.0833 ppb.
The RRF equals the ratio of the Control to the Baseline:
RRF = 88.5000 / 91.0833 = 0.972
Note that we report the RRF with three digits after the decimal point. The calculation of
the Baseline and Control averages does not involve any rounding or truncation.
9.4.1.3 RRF Calculation - Example 3
Assume the following default values:
RRF Setup:
Initial threshold value (ppb)
85
Minimum number of days in baseline at or above threshold
10
Minimum allowable threshold value (ppb)
70
Min number of days at or above minimum allowable threshold |~~
5
I- Enable Backstop minimum threshold for spatial fields
Backstop minimum threshold for spatial fields
Subrange first day of ozone season used in RRF |1
Subrange last day of ozone season used in RRF 1153
Pair days based on high concentration instead of date.
Spatial Gradient Setup:
Start Value
1
End Value
5
Assume that there are 15 days of data:
Abt Associates Inc.
264
June 2010
-------
Ozone Analysis: Details
Baseline day Baseline
value
1
84
2
85
3
85
4
82
5
78
6
76
7
70
8
70
9
70
10
67
11
64
12
63
13
62
14
62
15
59
Future day Future value
1
83
2
84
3
84
4
82
5
78
6
75
7
72
8
62
9
70
10
62
11
63
12
60
13
62
14
59
15
57
MATS will sort the data from high to low based on the Baseline values:
Baseline day Baseline Future day Future value
value
2
85
2
84
3
85
3
84
1
84
1
83
4
82
4
82
5
78
5
78
6
76
6
75
7
70
7
72
8
70
8
62
9
70
9
70
10
67
10
62
11
64
11
63
12
63
12
60
13
62
13
62
14
62
14
59
15
59
15
57
When you compare these sorted data with the Initial threshold value of 85 ppb, note that
there are only two (2) Baseline values at or above this threshold. Since there are fewer
days than the ten (10) specified as the Minimum number of days in baseline at or above
Abt Associates Inc.
265
June 2010
-------
Ozone Analysis: Details
threshold, MATS will then lower the threshold by one ppb to 84 ppb. There are three (3)
Baseline values at or above this lower threshold — still less than the value of ten (10) that
specified as the Minimum number of days in baseline at or above threshold. MATS will
lower the threshold again by one ppb, and eventually get to the Minimum allowable
threshold value of 70 ppb. At this point, there are still only nine (9) days at or above this
threshold.
MATS will take the nine days (highlighted in yellow) above the Minimum allowable
threshold value and then calculate separate averages for the Control and Baseline values:
Control average = 76.6667 ppb
Baseline average = 77.7778 ppb.
The RRF equals the ratio of the Control to the Baseline:
RRF = 76.6667 / 77.7778 = 0.986
Note that we report the RRF with three digits after the decimal point. The calculation of
the Baseline and Control averages does not involve any rounding or truncation.
9.4.1.4 RRF Calculation - Example 4
Assume the following default values:
RRF Setup:
Initial threshold value (ppb)
85
Minimum number of days in baseline at or above threshold
10
Minimum allowable threshold value (ppb)
70
Min number of days at or above minimum allowable threshold |~~
5
Enable Backstop minimum threshold for spatial fields
Backstop minimum threshold for spatial fields
Subrange first day of ozone season used in RRF |1
4
Subrange last day of ozone season used in RRF pj 53
4
I- Pair days based on high concentration instead of date.
Spatial Gradient Setup:
Start Value
1
End Value
5
Assume that there are 15 days of data:
Abt Associates Inc.
266
June 2010
-------
Ozone Analysis: Details
Baseline day Baseline Future day Future value
value
1
67
1
65
2
74
2
73
3
70
3
69
4
68
4
66
5
78
5
77
6
66
6
64
7
66
7
63
8
65
8
63
9
63
9
63
10
62
10
60
11
61
11
61
12
60
12
59
13
60
13
57
14
59
14
56
15
57
15
55
MATS will sort the data sorted from high to low based on the Baseline values:
Baseline day Baseline Future day Future value
value
5
78
5
77
2
74
2
73
3
70
3
69
4
68
4
66
1
67
1
65
6
66
6
64
7
66
7
63
8
65
8
63
9
63
9
63
10
62
10
60
11
61
11
61
12
60
12
59
13
60
13
57
14
59
14
56
15
57
15
55
When you compare these sorted data with the Initial threshold value of 85 ppb, note that
there are zero (0) Baseline values at or above this threshold. Since there are fewer days
than the ten (10) specified as the Minimum number of days in baseline at or above
Abt Associates Inc. June 2010
267
-------
Ozone Analysis: Details
threshold, MATS will lower the threshold by one ppb, and eventually get to the Minimum
allowable threshold value of 70 ppb.
At this point, there are still only three (3) days at or above this threshold — still less than
the Min number of days at or above minimum allowable. As a result, MATS will not
calculate a RRF and will set the future year design value to missing.
9.4.1.5 RRF Calculation - Example 5
Rather than assume the MATS default values, assume the Minimum allowable threshold
value is 60 ppb:
RRF Setup:
Initial threshold value (ppb)
85
Minimum number of days in baseline at or above threshold
10
Minimum allowable threshold value (ppb)
60
Min number of days at or above minimum allowable threshold |~~
5
I- Enable Backstop minimum threshold for spatial fields
Backstop minimum threshold for spatial fields
Subrange first day of ozone season used in RRF |1
Subrange last day of ozone season used in RRF 1153
Pair days based on high concentration instead of date.
Spatial Gradient Setup:
Start Value
1
End Value
5
Assume that there are 15 days of data:
Baseline day Baseline Future day Future value
value
1
67
1
65
2
74
2
73
3
70
3
69
4
68
4
66
5
78
5
77
6
67
6
64
7
66
7
63
8
65
8
63
9
63
9
63
Abt Associates Inc.
268
June 2010
-------
Ozone Analysis: Details
10
62
10
60
11
61
11
61
12
60
12
59
13
60
13
57
14
59
14
56
15
57
15
55
MATS will sort the data from high to low based on the Baseline values:
Baseline day Baseline Future day Future value
value
5
78
5
77
2
74
2
73
3
70
3
69
4
68
4
66
1
67
1
65
6
67
6
64
7
66
7
63
8
65
8
63
9
63
9
63
10
62
10
60
11
61
11
61
12
60
12
59
13
60
13
57
14
59
14
56
15
57
15
55
When you compare these sorted data with the Initial threshold value of 85 ppb, note that
there are zero (0) Baseline values at or above this threshold. Since there are fewer days
than the ten (10) specified as the Minimum number of days in baseline at or above
threshold, MATS will lower the threshold by one ppb, and eventually get to the Minimum
allowable threshold value of 62 ppb. At this point, there are ten (10) days at or above this
threshold.
MATS will take the ten days (highlighted in yellow) above the Minimum allowable
threshold value and then calculate separate averages for the Control and Baseline values:
Control average = 66.3000 ppb
Baseline average = 68.0000 ppb.
The RRF equals the ratio of the Control to the Baseline:
RRF = 66.3000 / 68.0000 = 0.975
Abt Associates Inc.
269
June 2010
-------
Ozone Analysis: Details
Note that we report the RRF with three digits after the decimal point. The calculation of
the Baseline and Control averages does not involve any rounding or truncation.
9.4.1.6 RRF Calculation Spatial Gradient with Backstop Threshold - Example 6
The following is an example showing the difference between RRF s calculated for Point
Estimates and RRFs calculated for Spatial Fields. The key difference is that MATS allows
you to choose a Backstop minimum threshold for spatial fields, which applies just to
Spatial Fields. This extra parameter allows you to calculate RRFs for Spatial Fields
exactly as you would for Point Estimates, except in the case when the minimum number of
days threshold cannot be met (MATS would return a -9 value for point estimates).
An example of such a case of where the two RRF calculations differ is when the number of
days at or above the Minimum allowable threshold value is less than the Minimum
number of days at or above minimum allowable threshold. In this case, MATS would
not calculate an RRF for a Point Estimate. However, if the Backstop minimum threshold
for spatial fields is set to some value lower than the Minimum allowable threshold
value, then MATS could potentially calculate an RRF for all or most grid cells in a Spatial
Field. The backstop minimum threshold allows the minimum threshold to be lowered to a
value below the Minimum allowable threshold value until the minimum number of days
is reached. The backstop threshold is only used for grid cells which do not have enough
days to meet the minimum number of days value with the minimum allowable threshold.
The backstop threshold does not change the calculation for grid cells that already meet the
minimum number of days.
RRF Setup:
Initial threshold value (ppb)
85
Minimum number of days in baseline at or above threshold
10
Minimum allowable threshold value (ppb)
70
Min number of days at or above minimum allowable threshold |~~
5
Enable Backstop minimum threshold for spatial fields
Backstop minimum threshold for spatial fields
60
Subrange first day of ozone season used in RRF |1
Subrange last day of ozone season used in RRF 1153
Pair days based on high concentration instead of date.
Spatial Gradient Setup:
Start Value
1
End Value
5
Abt Associates Inc.
270
June 2010
-------
Ozone Analysis: Details
Assume that there are 15 days of data for a grid cell:
Baseline day Baseline Future day Future value
value
1
67
1
65
2
74
2
73
3
70
3
69
4
68
4
66
5
78
5
77
6
67
6
64
7
66
7
63
8
65
8
63
9
63
9
63
10
62
11
61
11
61
10
60
12
60
12
59
13
60
13
57
14
59
14
56
15
57
15
55
MATS will sort the data from high to low based on the Baseline values:
Baseline day Baseline Future day Future value
value
5
78
5
77
2
74
2
73
3
70
3
69
4
68
4
66
1
67
1
65
6
67
6
64
7
66
7
63
8
65
8
63
9
63
9
63
10
62
11
61
11
61
10
60
12
60
12
59
13
60
13
57
14
59
14
56
15
57
15
55
When you compare these sorted data with the Initial threshold value of 85 ppb, note that
there are zero (0) Baseline values at or above this threshold. Since there are fewer days
Abt Associates Inc.
271
June 2010
-------
Ozone Analysis: Details
than the ten (10) specified as the Minimum number of days in baseline at or above
threshold, MATS will lower the threshold by one ppb, and eventually get to the Minimum
allowable threshold value of 70 ppb.
At this point, there are still only three (3) Baseline values (highlighted in yellow) at or
above this lower threshold — still less than the value of ten (10) that specified as the
Minimum number of days in baseline at or above threshold. This fails the test for
calculating an RRF for a Point Estimate. However, there is still a possibility that MATS
can calculate an RRF for a Spatial Field. MATS just needs to find at least five values
MATS will lower the threshold again by one ppb. At a threshold of 68 ppb, there are four
(4) days. MATS will lower the threshold again by one ppb. At a value of 67 ppb, there are
six (6) days at or above the Backstop minimum threshold for spatial fields.
Baseline day Baseline Future day Future value
value
5
78
5
77
2
74
2
73
3
70
3
69
4
68
4
66
1
67
1
65
6
67
6
64
7
66
7
63
8
65
8
63
9
63
9
63
10
62
11
61
11
61
10
60
12
60
12
59
13
60
13
57
14
59
14
56
15
57
15
55
Since MATS is looking for at least five days, MATS will take these six days (highlighted
in yellow) above the Backstop minimum threshold for spatial fields and then calculate
separate averages for the Control and Baseline values:
Control average = 70.0000 ppb
Baseline average = 71.4000 ppb.
The RRF equals the ratio of the Control to the Baseline:
RRF = 70.0000 / 71.4000 = 0.980
Note that we report the RRF with three digits after the decimal point. The calculation of
the Baseline and Control averages does not involve any rounding or truncation.
Abt Associates Inc.
272
June 2010
-------
Ozone Analysis: Details
9.4.2 Spatial Gradient Setup
In using a spatial gradient to estimate ozone levels, MATS estimates ozone levels in
unmonitored locations by using the values of a nearby monitored data scaled by a ratio of
model values. The ratio, or spatial gradient, is a mean of model values at the unmonitored
location over the mean of the model values at a monitor.
Note that several "nearby" monitors (and their associated model values) are used in the
calculation of ozone values at an unmonitored location. MATS uses a process called
Voronoi Neighbor Averaging (VNA) to identify these neighbors, and then takes an inverse
distance-weighted average of these monitors.
MATS sorts the daily 8-hour maximum ozone values from high to low, averages a certain
number of these values (by default the top five), and then uses these averages in the
calculation of the spatial gradient. Note that the highest days for Cell A and Cell E are
determined independently of each other.
If you want to use a different set of days for the gradient adjustment, you can do so with
the Start Value and End Value. MATS assigns an index of value of 1 to the highest daily
8-hour maximum ozone value in each grid cell. The second-highest an index value of 2.
And so on. Using the Start Value and the End Value, you
can identify the values that you
want to average by using this index.
Spatial Gradient Calculation - Example 1
Assume a Start Value of "1" and an End Value of "5":
RRF Setup:
Initial threshold value (ppb)
85
Minimum number of days in baseline at or above threshold
10
Minimum allowable threshold value (ppb)
70
Min number of days at or above minimum allowable threshold |~~
5
I- Enable Backstop minimum threshold for spatial fields
Backstop minimum threshold for spatial fields
Subrange first day of ozone season used in RRF |1
Subrange last day of ozone season used in RRF 1153
Pair days based on high concentration instead of date.
Spatial Gradient Setup:
Start Value
1
End Value
5
Abt Associates Inc.
273
June 2010
-------
Ozone Analysis: Details
For this example calculation, assume that we have one monitor and we want to use this
monitor to estimate the ozone level at the center of a nearby grid cell. Further assume that
the monitor resides in grid cell "A" and the we want to estimate the ozone level in grid cell
"E".
With the default Start Value equal to one (1) and the default End Value, equal to five (5),
MATS will average the five highest daily 8-hour maximum ozone values. Note, however,
that the highest days for Cell A and Cell E are determined independently of each other.
Assume that there are 15 days of data:
Day
Cell A
Day
Cell E
1
100
1
68
2
100
2
73
3
98
3
74
4
97
4
78
5
95
5
72
6
95
6
69
7
90
7
77
8
85
8
63
9
84
9
65
10
83
10
61
11
83
11
60
12
83
12
62
13
79
13
58
14
78
14
56
15
78
15
57
MATS will sort the data for cell A from high to low. Independently, MATS will also sort
the data for Cell E from high to low. In this example, Day 1 has the highest value for cell
A, while the highest value for cell E falls on Day 2.
Day
Cell A
Day
Cell E
1
100
4
78
2
100
7
77
3
98
3
74
4
97
2
73
5
95
5
72
6
95
6
69
7
90
1
68
8
85
9
65
9
84
8
63
10
83
12
62
11
83
10
61
Abt Associates Inc.
274
June 2010
-------
Ozone Analysis: Details
12 83 11 60
13 79 13 58
14 78 15 57
15 78 14 56
MATS will take the top 5 days (highlighted in yellow) and then calculate separate averages
for cell A and cell E:
cell E average = 74.8000 ppb
cell A average = 98.0000 ppb.
The Spatial Gradient equals the ratio of Cell E to Cell A:
Spatial Gradient = 74.8000 / 98.0000 = 0.763
Note that we report the Spatial Gradient with three digits after the decimal point. The
calculation of the averages does not involve any rounding or truncation.
9.4.2.2 Spatial Gradient Calculation - Example 2
Assume a Start Value of "2" and an End Value of "3":
RRF Setup:
Initial threshold value (ppb)
85
Minimum number of days in baseline at or above threshold
10
Minimum allowable threshold value (ppb)
70
Min number of days at or above minimum allowable threshold |~~
5
Enable Backstop minimum threshold for spatial fields
Backstop minimum threshold for spatial fields
60
Subrange first day of ozone season used in RRF |1
Subrange last day of ozone season used in RRF 1153
Pair days based on high concentration instead of date.
Spatial Gradient Setup:
Start Value
2
End Value
3
For this example calculation, assume that we have one monitor and we want to use this
monitor to estimate the ozone level at the center of a nearby grid cell. Further assume that
Abt Associates Inc.
275
June 2010
-------
Ozone Analysis: Details
the monitor resides in grid cell "A" and the we want to estimate the ozone level in grid cell
"E"
With the default Start Value equal to two (2) and the default End Value, equal to three (3),
MATS will average the second and third highest daily 8-hour maximum ozone values.
Assume that there are 15 days of data:
Day
Cell A
Day
Cell E
1
100
1
68
2
100
2
73
3
98
3
74
4
97
4
78
5
95
5
72
6
95
6
69
7
90
7
77
8
85
8
63
9
84
9
65
10
83
10
61
11
83
11
60
12
83
12
62
13
79
13
58
14
78
14
56
15
78
15
57
MATS will sort the data for cell A from high to low. Independently, MATS will also sort
the data for Cell E from high to low.
Day
Cell A
Day
Cell E
1
100
4
78
2
100
7
77
3
98
3
74
4
97
2
73
5
95
5
72
6
95
6
69
7
90
1
68
8
85
9
65
9
84
8
63
10
83
12
62
11
83
10
61
12
83
11
60
13
79
13
58
14
78
15
57
15
78
14
56
Abt Associates Inc.
276
June 2010
-------
Ozone Analysis: Details
MATS will take the second and third highest days (highlighted in yellow) and then
calculate separate averages for cell A and cell E:
cell E average = 75.5000 ppb
cell A average = 99.0000 ppb.
The Spatial Gradient equals the ratio of Cell E to Cell A:
Spatial Gradient = 75.5000 / 99.0000 = 0.763
Note that we report the Spatial Gradient with three digits after the decimal point. The
calculation of the averages does not involve any rounding or truncation.
9.4.2.3 Spatial Gradient Calculation - Example 3
Assume a Start Value of "4" and an End Value of "4":
RRF Setup:
Initial threshold value (ppb)
85
Minimum number of days in baseline at or above threshold
10
Minimum allowable threshold value (ppb)
70
Min number of days at or above minimum allowable threshold |~~
5
Enable Backstop minimum threshold for spatial fields
Backstop minimum threshold for spatial fields
60
Subrange first day of ozone season used in RRF |1
Subrange last day of ozone season used in RRF 1153
Pair days based on high concentration instead of date.
Spatial Gradient Setup:
Start Value
4
End Value
4
For this example calculation, assume that we have one monitor and we want to use this
monitor to estimate the ozone level at the center of a nearby grid cell. Further assume that
the monitor resides in grid cell "A" and the we want to estimate the ozone level in grid cell
ME".
With the default Start Value equal to four (4) and the default End Value, equal to four (4),
MATS will only use the fourth highest daily 8-hour maximum ozone value in each grid
cell.
Abt Associates Inc.
277
June 2010
-------
Ozone Analysis: Details
Assume that there are 15 days of data:
Day
Cell A
Day
Cell E
1
100
1
68
2
100
2
73
3
98
3
74
4
97
4
78
5
95
5
72
6
95
6
69
7
90
7
77
8
85
8
63
9
84
9
65
10
83
10
61
11
83
11
60
12
83
12
62
13
79
13
58
14
78
14
56
15
78
15
57
MATS will sort the data for cell A from high to low. Independently, MATS will also sort
the data for Cell E from high to low.
Day
Cell A
Day
Cell E
1
100
4
78
2
100
7
77
3
98
3
74
4
97
2
73
5
95
5
72
6
95
6
69
7
90
1
68
8
85
9
65
9
84
8
63
10
83
12
62
11
83
10
61
12
83
11
60
13
79
13
58
14
78
15
57
15
78
14
56
MATS will take the fourth highest day (highlighted in yellow) for cell A and cell E:
cell E = 73 ppb
cell A = 97 ppb.
Abt Associates Inc.
278
June 2010
-------
Ozone Analysis: Details
The Spatial Gradient equals the ratio of Cell E to Cell A:
Spatial Gradient = 73 / 97 = 0.753
Note that we report the Spatial Gradient with three digits after the decimal point.
9.5 Final Check
The Final Check window verifies the selections that you have made.
Click the button Press here to verify your selections. If there are any errors, MATS will
present a message letting you know. For example, if the path to a model file is invalid —
perhaps you misspelled the file name — you would get the following error:
Verily inputs
Press here to verily your selections...
Checking...
A valid monitor data file, base and future model files are required.
Base year data file missing.
Abt Associates Inc.
279
June 2010
-------
Ozone Analysis: Details
After making the necessary correction, click the button Press here to verify your
selections. .
- Verily inputs
Press here to verify your selections... |
Checking...
Check OK. Press the finish button to continue..
When your window looks like the window above, click either Save Scenario & Run or
Save Scenario. Save Scenario & Run will cause MATS to immediately run the scenario.
9.5.1 Running MATS in Batch Mode
The Save Scenario button will save the scenario as a configuration file (.cfg file). The
"*.cfg" file will be saved in the .\MATS\output directory. Several .cfg files can be created
with the MATS interface and run later in batch mode. To do this, edit the default batch file
located in the .\MATS directory. The file "batchmats.bat" should be edited with a text
editor to point to the name and location of the .cfg files that will be run in batch mode.
[C:\Documents and Settings\btimin\Desktop\MATS-5-08\batchmats.bat] [_ |(n|(x|
|^J File Edit Search Project View Format Column Macro Advanced Window Help
D a? ¦ h|
# ^ S ^ I w, | H ieseis 14 wt iP gs
batchmats.bat
! 1
start /wait
HATS.exe "c:\PragraKi Files\Abt. Associates\MATS\\output \default.cfg" Tj
start /wait
MATS.exe "c:\Pi:Ggrsm Files\Abt AssQciates\ MATS\ \ output\def ault2 . cfg"
l .1^
«l
For Help, press F1 t
.n I, Col, 1, CW DOS | Mod: 5/23/2008 5:02:49PM |File Size: 163 |lN5 f
After editing the batchmats.bat file, simply run the .bat file. MATS will start and run in
the background.
Abt Associates Inc.
280
June 2010
-------
Visibility Analysis: Quick Start Tutorial
10 Visibility Analysis: Quick Start Tutorial
In this tutorial you will forecast visibility levels at Class I Areas in the United States. The
steps in this analysis are as follows:
Step 1. Start MATS. Start the MATS program and choose to do a Visibility analysis.
Step 2. Output Choice. Choose the output to generate. In this example, you will
forecast visibility levels using the new IMPROVE algorithm and model data at the
IMPROVE monitors.
Step 3. Data Input. Choose the data files for input to MATS.
Step 4. Filtering. Choose the years of monitor and model data that you want to use, and
then choose the particular monitors in these data that you want to include in the analysis.
Step 5. Final Check. Verify the choices you have made.
Step 6. Load & Map Results. Load your results and prepare maps of your forecasts.
Step 7. Working with Configuration File. Examine the Configuration file that stores the
choices that you made underlying your analysis.
Each step is explained in detail below. Additional details are provided in the section
Visibility Analysis: Details.
10.1 Stepl. Start MATS
Double-click on the MATS icon on your desktop, and the following window will appear:
Abt Associates Inc.
281
June 2010
-------
Visibility Analysis: Quick Start Tutorial
Umats
Help T
Start Map View Output Navigator
Ozone Analysis
Visibility Analysis
Annual PM Analysis
Daily PM Analysis
Stop Info
Click the Visibility Analysis button on the main MATS window. This will bring up the
Configuration Management window.
Configuration Management
(* plreate New Configuration!
C Open Existing Configuration
Go
Cancel
A Configuration allows you to keep track of the choices that you make when using MATS.
For example, after generating results in MATS, you can go back, change one of your
choices, rerun your analysis, and then see the impact of this change without having to enter
in all of your other choices. For this example, we will start with a new Configuration.
Choose Create New Configuration and click the Go button. This will bring up the
Choose Desired Output window.
10.2 Step 2. Output Choice
The Choose Desired Output window allows you to choose the output that you would like
Abt Associates Inc.
282
June 2010
-------
Visibility Analysis: Quick Start Tutorial
to generate. MATS allows you to calculate future year (forecast) visibility levels at Class I
Areas.
In the Scenario Name box type "Tutorial Visibility" - this will be used to keep track of
where your results are stored and the variable names used in your results files. Leave the
box checked next to Temporally-adjust visibility levels at Class I Areas. MATS will
create forecasts for each Class I Area in your modeling domain.
MATS provides two algorithms for calculating visibility — an "old version" and a "new
version" of the IMPROVE visibility algorithm (see IMPROVE 2006) (The old and new
versions are discussed in Desired Output section of the Visibility Analysis: Details
chapter.) Choose the new version.
A single IMPROVE monitor is associated with each Class I Area. MATS multiplies the
monitor value with a relative response factor (RRFl which is the the modeled future-year
visibility divided the modeled current-year visibility. In calculating the RRF, MATS
allows you to use either the model values in the grid cell at the IMPROVE monitor or to
use the model values in the grid cell at the Class I Area centroid. Choose the default option
of using model values in the grid cell at the monitor. (For additional details see the
Desired Output section of the Visibility Analysis: Details chapter.)
Check the box next to Automatically extract all selected output files. MATS will create a
separate folder called "Tutorial Visibility" in the MATS "Output" folder, and then export
.CSV files with the results of your analysis. Alternatively, you can export the results from
the Output Navigator, but checking this box is a little easier.
Abt Associates Inc.
283
June 2010
-------
Visibility Analysis: Quick Start Tutorial
Choose Desired Output
Point Estimates
Scenario Name : |Tutorial Visibility
Forecast
p T emporally-adjust visibility levels at Class 1 Areas
IMPROVE Algorithm
C use old version
(• use new version
© Use model grid cells at monitor
Use model grid cells at Class 1 area centroid
Actions on run completion
p Automatically extract all selected output files
Next >
Cancel
When your window looks like the window above, click Next. This will bring up the Data
Input window.
The Data Input window allows you to choose the monitor data and the model data that
you want to use. As discussed in more detail in the Visibility Analysis: Details chapter,
MATS calculates the ratio of the model data to calculate a relative response factor fRRF)
for the 20% best and 20% worst visibility days separately. MATS then multiplies the
visibility level measured at the monitor for the best days with the RRF for the best days to
calculate a future-year estimate for visibility on the best visibility days. MATS performs
an analogous calculation for the worst visibility days.
MATS comes loaded with IMPROVE visibility monitor values from 2000 through 2004.
It also comes loaded with an example model output dataset for visibility for 2001 and
2015. These are the key ingredients for creating your visibility forecasts.
Use the default settings in the Data Input window. The window should look like the
following:
10.3 Step 3. Data Input
Abt Associates Inc.
June 2010
284
-------
Visibility Analysis: Quick Start Tutorial
¦ Choose Desired Output
Data Input
¦ Filtering
ESI Final Check
Data Input
Monitor Data
IMPROVE Monitor Data • Old Algorithm JCAProgrann Files^Abt Associates\MATS\Sarnp ¦¦¦
IMPROVE Monitor Data ¦ New Algorithm |006-daily IMPROVE-all data-new equation.csv| ¦¦¦!
Model Data
Baseline File
Forecast File
Using Model Data
T emporal adjustment at monitor
|S\SampleData\2002cc_EUS_PM25_sub.csv
|S\SampleData\2020cc_ELIS_PM25_sub.csv
3x3 "r
< Back
Next >
Cancel
Note that MATS gives you the option to use model data in different ways when calculating
forecasts at each monitor. The example model datasets are at 36km resolution. Therefore,
the default is to use a 3x3 array of model cells around each monitor. This is described in
more detail in the Using Model Data section of the Visibility Analysis: Details chapter.
When your window looks like the window above, click Next. This brings up the visibility
Filtering window.
10.4 Step 4. Filtering
The Filtering window has two sets of functions. The first involves identifying the years of
monitor and model data that you want to use. The second involves identifying the
particular monitors in these data that you want to include in the analysis. Use the default
settings pictured in the screenshot below.
Choose Visibility Data Years
Specify the range of visibility monitor data that you want to use. The default is to use all
of the available data: 2000 through 2004. (That is, Start Monitor Year set to 2000 and
End Monitor Year set to 2004.)
Choose the Base Model Year. This should match the metorological year that is being
modeled. It should fall within the range specified by the Start Monitor Year and the End
Abt Associates Inc.
285
June 2010
-------
Visibility Analysis: Quick Start Tutorial
Monitor Year. The Base Model Year for the example dataset is 2001.
Valid Visibility Monitors
Identify the monitors that you want to include in the analysis. First, specify the
Minimum years required for a valid monitor. MATS excludes from the analysis any
monitors with fewer than the Minimum years required for a valid monitor. The
default value is 3 years.
Specify the Maximum Distance from Domain [km]. Monitors that are further than the
Maximum Distance from Domain [km] are excluded from the analysis. The default
value is 25 kilometers (km).
m
I Choose Desired Output
¦ Data Input
Filtering
I Final Check
Filtering
Choose Visibility Data Years
Start Monitor Year End Monitor Year Base Model Year
ISNHflQ 12004 T] 12002 V]
Valid Visibility Monitors
Minimum years required for a valid monitor 3
Man Distance from Centroid to Gridcell Center [km] 25
< Back
Next >
Cancel
10.5 Step 5. Final_Check
The Final Check window verifies the choices that you have made. For example, it makes
sure that the paths specified to each of the files used in your Configuration are valid.
Click on the Press here to verify selections button.
Abt Associates Inc.
286
June 2010
-------
Visibility Analysis: Quick Start Tutorial
Choose Desired Output
Data Input
Filtering
Final Check
Final Check
Verify inputs
Press here to verity your selections..
Checking...
Check OK. Press the finish button to continue..
Save Scenario
< Back Save Scenario Si Run
Cancel
If you encounter any errors, go back to the choices you have previously made by clicking
on the appropriate part (e.g., Data Input) of the tree in the left panel, and then make any
changes required.
When your window looks like the window above, click either Save Scenario & Run or
Save Scenario. Save Scenario & Run will cause MATS to immediately run the scenario.
A temporary, new Running tab will appear (in addition to the Start, Map View and
Output Navigator tabs).
Abt Associates Inc.
287
June 2010
-------
Visibility Analysis: Quick Start Tutorial
r —
¦
*
MATS
(ZB>
3
Help T
I Start Map View Output Navigator Running
t
Close
Name
Last Message
Tutorial Visibility.asr
Starting Visibility Analysis
Note that MATS is very computation-intensive, so if you try to work with other programs
in addition they may run very slowly. When the calculations are complete, a small window
indicating the results are Done will appear. Click OK.
Done
Mill
After clicking OK, the Output Navigator tab will be active. (The Running tab will no
longer be seen.). MATS will automatically load the output files associated with the .asr
configuration that just finished running.
Abt Associates Inc.
288
June 2010
-------
Visibility Analysis: Quick Start Tutorial
011®
Help
Start Map View | Output Navigator I
Load
Extract All
Highlight file of interest and right-click to view options to Map, View, and Extract the data.
Name
i|Type
| Size
B Configuration/Log Files
j- Configuration
Configuration
54kb
s- Log File
Run Log
Okb
S Output Files
h Tutorial Visibility- Class 1 Area and IMPROVE Monitor Identifiers and Locations
Monitor Network
17kb
; Tutorial Visibility - Forecasted Visibility - all design values
Monitor Network
6kb
f Tutorial Visibility - Forecasted Visibility Data
Monitor Network
2kb
j- Tutorial Visibility - Used Model Reference Cells - Base Data
Monitor Network
35kb
Tutorial Visibility-Used Model Reference Cells - Future Data
Monitor Network
35kb
Stop Info
The next step (click here) shows you how to map your results with the Output Navigator.
For more details on mapping and other aspects of the Output Navigator, there is a
separate chapter on the Output Navigator.
10.6 Step 6. Load and Map Results
After generating your results, Output Navigator can be used to load and/or map them. If a
run just finished, the output files will already be loaded into output navigator.
If files from a previous run need to be loaded then click on the Load button and choose
the Tutorial Visibility.asr file.
Abt Associates Inc.
289
June 2010
-------
Visibility Analysis: Quick Start Tutorial
HP®
Help -
Start Map View |Output Navigator l|
Load
Name [Type [Size
No file loaded
]| Stop Info
Click on the Load button and choose the Tutorial Visibility.asr file.
Abt Associates Inc. June 2010
290
-------
Visibility Analysis: Quick Start Tutorial
Select a Configuration
My Recent
Documents
a
Desktop
My Documents
My Computer
My Network
Places
Look in: output
Example 03
jTutorial 03
jTutorial Visibility
Example 03.asr
Ml Tutorial 03.asr
Tutorial Visibility.asr
3
CS &
File name:
Files of type:
[Tutorial Visibility, asr
73
Open
| ASR files ("asr)
73
Cancel
Under Configuration/Log Files, you will see two files:
Configuration: keeps track of the assumptions that you have made in your analysis.
Log File : provides information on a variety of technical aspects regarding how a results
file (*.ASR) was created.
Under Output Files you will see:
Tutorial Visibility - Forecasted Visibility - all design values: baseline and forecasted
visibility levels for the best and worst days for each year of the five year base period.
Tutorial Visibility - Used Model Grid Cells - Base Data: baseline model values for PM
species for the grid cells and days used in the RRF calculations.
Tutorial Visibility - Used Model Grid Cells - Future Data: future-year model values for
PM species for the grid cells and days used in the RRF calculations.
Tutorial Visibility - Forecasted Visibility Data: baseline and forecasted deciview values
for the best and worst days (averaged across up to five years). Also includes
species-specific relative response factors for the best and worst days.
Tutorial Visibility - Class 1 Area and IMPROVE Monitor Identifiers and Locations:
monitor latitude and longitude.
Abt Associates Inc.
291
June 2010
-------
Visibility Analysis: Quick Start Tutorial
011®
Help
Start Map View | Output Navigator I
Load
Extract All
Highlight file of interest and right-click to view options to Map, View, and Extract the data.
Name
i|Type
| Size
B Configuration/Log Files
j- Configuration
Configuration
54kb
s- Log File
Run Log
Okb
S Output Files
Tutor
al Visibility - Class 1 Area and IMPROVE Monitor Identifiers and Locations
Monitor Network
17kb
• Tutor
al Visibility - Forecasted Visibility - all design values
Monitor Network
6kb
j™Tutor
al Visibility - Forecasted Visibility Data
Monitor Network
2kb
h Tutor
al Visibility- Used Model Reference Cells - Base Data
Monitor Network
35kb
Tutor
al Visibility - Used Model Reference Cells - Future Data
Monitor Network
35kb
1 Stop Info
Right-click on the file Tutorial Visibility - Forecasted Visibility Data. This gives you three
options: Add to Map, View, and Extract. Choose the Add to Map option.
Abt Associates Inc.
292
June 2010
-------
Visibility Analysis: Quick Start Tutorial
Help T
Start Map View | Output Navigator I
Load
Extract All
Highlight file of interest and right-click to view options to Map, View, and Extract the data.
|Type
Name
Size
- Configuration/Log Files
j- Configuration
Log File
S Output Files
Tutorial Visibility - Forecasted Visibility - all design values
Tutorial Visibility - Used Model Reference Cells - Base Data
|-Tutorial Visibility - Used Model Reference Cells - Future Data
Tutorial Visibility - Forecasted Visp
Add To Map
Configuration
Run Log
54kb
Okb
Monitor Network 6kb
Monitor Network 35kb
Monitor Network 35kb
Monitor Network 2kb
Tutorial Visibility - Class 1 Area ar yjew
Extract
itor Identifiers and Locations Monitor Network 17kb
Stop Info
This will bring up the Map View.
Abt Associates Inc.
293
June 2010
-------
Visibility Analysis: Quick Start Tutorial
Help
Start | Map View | Output Navigator
" ^ ^ Standard Layers'
J Data Loaded ;|
@ • Tutorial Visibility - Forecasted"
$ is-
Long: -194.92635, Lat: 34.91279 ***
Extent: Min(-159.661,2.334) Max(-&"*H51.197)
To view an enlarged map, use the Zoom to an area Task Bar button on the far left.
Choose the Continental US.
Maryland
New England
Southern California
Texas
Washington DC
:o re caste d'
Start | Map View | Output Navigator
Long: -189.22724, Lat: 38.81900 ***
Extent: Min(-159.661,2.334) Max(-E***l4,51.197)
Full Extent
N S3
Standard Layers T
Edit Zoom Frames
Add Current View to List
Abt Associates Inc.
294
June 2010
-------
Visibility Analysis: Quick Start Tutorial
The map will then zoom to the Continental US.
Long:-126.33561, Lat: 39.01791
Extent: Min(-122.607,21.608) Max(***.820,45.788)
Start | Map View | Output Navigator
@ • Tutorial Visibility-Forecasted'
+v' + ^ ^ T ^ Standard Layers "
:Data Loaded
Stop- Info
To more easily view the location of monitors in particular states, uncheck US Counties
using the Standard Layers drop down menu on the far right of the Task Bar. Your
window should look like the following:
Abt Associates Inc.
295
June 2010
-------
Visibility Analysis: Quick Start Tutorial
JiData Loaded |
0 • Tutorial Visibility-Forecasted
Standard Layers *
¦/ US States
an
Start | Map View | Output Navigator
Long:-110.31785, Lat: 52.89023
Extent: Min(-122.607,21.608) Max(-*°**.820,45.788)
Right click on the "Tutorial Visibility - Forecasted Visibility Data" layer in the panel on
the left side of the window. Choose the Plot Value option.
I Help ^
Long: -121.77797, Lat: 22.45886 ***
Extent: Min(-122.607,21.608) Max(-***.820,45.788)
Start | Map View | Output Navigator
JiData Loaded ||
m
Standard Layers »
@ • Tutorial Visibilih'-F|-"'ar'«etQ^'
Remove
Export as CSV File
Abt Associates Inc.
296
June 2010
-------
Visibility Analysis: Quick Start Tutorial
This will bring up Shape Class Breaks window. In the Value drop-down list, choose the
variable "dv best" — this is forecasted visibility design value for the best visibility days in
2015. (Note that the Date box defaults to the baseline year; in this case 2002)
Shape Class Breaks
Layer Name: Tutorial Visibility - Forecasted Visibility Data
Value:
Date
dv best
2002
(* Bins C Unique Values
Class Count: [5 ^ Marker Sizing: [o ^
Start Color
End Color
iJS Clear Breaks V' Apply X Close
Click Apply and then click Close. This will bring you back to the Map View window.
jg MATS
Help
Long:-113.19533, Lat: 23.48534 "¦
Extent: Min(-122.607,21.608) Max(^
!.820,45.788)
Start | Map View | Output Navigator
+ ' + ~. Ky K|| ^ Standard Layers "
JiData Loaded ;|
@ ° Tutorial Visibility- Forecasted
© • dv.best 8.94 to 10.62
@ • dv_best 10.62 to 12.03
@ ° dv_best 12.62 to 14.01
@ ° dv_best 14.74 to 14.74
Stop
Examine the other variables:
Abt Associates Inc.
297
June 2010
-------
Visibility Analysis: Quick Start Tutorial
dv worst: forecasted deciview values for 20% worst days;
dvbest, forecasted deciview values for 20% best days
base best, baseline design values for best days;
base worst baseline design values for worst days;
rrf b crustal, rrf bno3, rrf b oc, rrf b ee, rrf b cm, and rrf b so4\ relative
response factor used to forecast the best visibility days;
rrf it' crustal, rrf it' no3, rrf it' oc, rrf it' ec, rrf it' cm, and rrf^w so4\ relative
response factor used to forecast the worst visibility days;
This is just a brief summary of the mapping possibilities available. For more details, there
is a separate chapter on the Map View.
10.7 Step 7. Working with Configuration File
Configurations keep track of the choices that you have made in your analysis. There are
two ways that you can access your configuration. First, you can view your configuration
using the Output Navigator. Right-click on Configuration and choose View.
Abt Associates Inc.
298
June 2010
-------
Visibility Analysis: Quick Start Tutorial
This will take you to the Tutorial Visibility configuration that you used to generate your
visibility results.
Bl Choose Desired Out|
¦ Data Input
Choose Desired Output
Final Check
n—i
Point Estimates
Scenario Name
Forecast
|7 T ernporally-adjust visibility levels at Class 1 Areas
IMPROVE Algorithm
C use old version
(• use new version
© Use model grid cells at monitor
Use model grid cells at Class 1 area centroid
Actions on run completion
|7 Automatically extract all selected output files
< Back
Next >
Cancel
A second way to access your Tutorial Visibility configuration is to go back to the Start
window.
Abt Associates Inc.
299
June 2010
-------
Visibility Analysis: Quick Start Tutorial
Umats
Help
Start
Map View Output Navigator
Ozone Analysis
Visibility Analysis
Annual PM Analysis
Daily PM Analysis
Stop Info
Click on the Visibility Analysis button. This will bring up the Configuration
Management window.
Configuration Management
(* jCreate New Configuration!
r Open Existing Configuration
Go
Cancel
Choose Open Existing Configuration and then click the Go button. This will bring up
the Select a Configuration window. Find the Tutorial Visibility.asr file that you
generated.
Abt Associates Inc.
300
June 2010
-------
Visibility Analysis: Quick Start Tutorial
Select a Configuration
Look in:
My Recent
Documents
J
Desktop
Documents
Si
My Computer
Mj,1 Network
Places
[ET output
Example 03
Tutorial 03
I Tutor ial Visibility
HExample 03.asr
E^l Tutorial 03.asr
Tutorial Visibility.asr
?j\x\
"3
© &
File name:
Tutorial Visibility.asr
"3
Open
Files of type:
[asr files (".asr)
3
Cancel
Click Open and this will bring you your Tutorial Visibility configuration. Choose the Use
model grid cells at Class 1 area centroid option.
Abt Associates Inc.
301
June 2010
-------
Visibility Analysis: Quick Start Tutorial
Choose Desired Oi
¦ Data Input
¦ Filtering
IBsl Final Check
Choose Desired Output
Point Estimates
Scenario Name : |Tutorial Visibility
Forecast
p T emporally-adjust visibility levels at Class 1 Areas
IMPROVE Algorithm
C use old version (• use new version
Use model grid cells at monitor
© lUse model grid cells at Class 1 area centroicj
Actions on run completion
|7 Automatically extract all selected output files
< Back
Next >
Cancel
MATS will now calculate RRFs using model data located over the center of each Class 1
area, instead of using model data located over the monitor linked to each Class 1 area.
To reflect this change in your analysis, change the Scenario Name box to Tutorial
Visibility - Model at Class 1.
Abt Associates Inc.
302
June 2010
-------
Visibility Analysis: Quick Start Tutorial
Choose Desired Output
Final Check
Point Estimates
Scenario Name : [Tutorial Visibility - Model at Class I
Forecast
|7 T emporally-adjust visibility levels at Class 1 Areas
rIMPROVE Algorithm
C use old version
(*
use new version
Use model grid cells at monitor
® Use model grid cells at Class 1 area centroid
Actions on run completion
|7 Automatically extract all selected output files
< Back
Next >
Cancel
Keep all of the other assumptions the same. At the Filtering window, click the Finish
button. MATS will generate a new set of results and save them in a file called: Tutorial
Visibility - Model at Class l.asr. You can then view and map your results in the same way
as with other results files.
Abt Associates Inc.
303
June 2010
-------
Visibility Analysis: Details
11 Visibility Analysis: Details
MATS can calculate baseline and future-year visibility levels for the best and worst days
for Class I Areas — these estimates are referred to as Point Estimates. as they refer to
particular locations. MATS gives you several options for how to generate these estimates,
and keeps track of the choices you make with a Configuration.
When you begin the process of generating visibility estimates, MATS provides an option
to start a new Configuration or to open an existing Configuration.
Configuration Management
(5" iCreate New Configuration!
C Open Existing Configuration
Go
Cancel
Select your option and then click Go.
MATS will then step you through a series of windows with choices for your analysis.
Choose Desired Output. Choose whether you want to calculate Point Estimates at
IMPROVE monitors or at Class I Area centroids and whether to use the old or new
version of the IMPROVE visibility equation.
Data Input. Specify the air modeling and monitoring data that you want to use. Specify
which model grid cells will be used when calculating RRFs at monitor locations.
Filtering. Choose the years of monitoring data. Identify valid monitors.
MATS comes with a set of default choices and an example set of input files. If desired you
can use these defaults and skip to the Final Check window and click the Finish button to
generate your calculations.
11.1 Choose Desired Output
In the Choose Desired Output window, you specify the Scenario Name that you would
like to use, as well as choices regarding how you would like to calculate future year
(forecast) visibility levels for Class I Areas. As discussed in the section on Forecasting
Visibility, the forecast calculations have a number of steps. At the end of this section,
there is an example of these calculations.
You may use the "old version" or "new version" of the IMPROVE Equation (IMPROVE.
2006). which MATS uses to translate PM levels (measured in ug/m3) to visibility levels
(measured in extinction or deciviews). You may also choose between using model data at
Abt Associates Inc. June 2010
304
-------
Visibility Analysis: Details
the monitor or model data at the center of the Class I Area.*
By checking the box next to Automatically extract all selected output files, MATS will
create a separate folder with your chosen Scenario Name in the MATS "Output" folder,
and then export .CSV files with the results of your analysis. Alternatively, you can export
the results from the Output Navigator, but checking this box is a little easier.
CI Choose Desired Out|
¦ Data Input
¦ Filtering
Choose Desired Output
Final Check
Point Estimates
Scenario Name : |EKannple Visibility
Forecast
|7 T emporally-adjust visibility levels at Class 1 Areas
rIMPROVE Algorithm
C use old version
C
use new version
© Use model grid cells at monitor
Use model grid cells at Class 1 area centroid
Actions on run completion
^Automatically extract all selected output files!
< Back
Next >
Cancel
* Monitors assigned to represent a Class I Area are generally close to the Class I Area. However, in
some cases, the distance can be substantial. For example, the YELL2 monitor in Wyoming (44.5653
latitude, -110.4002 longitude) is located more than a degree longitude away from the Red Rocks Lake
Class I Area (44.64 latitude, -111.78 longitude). By default, MATS uses model data at the monitor.
11.1.1 Scenario Name
The Scenario Name allows you to uniquely identify each analysis that you conduct. It is
used in three ways.
(1) Results file name. The results file is given the Scenario Name (e.g., Tutorial
Visibility, asr). Note that the extension (ASR) is specifically designated just for MATS
and can only be used by MATS.
Abt Associates Inc.
305
June 2010
-------
Visibility Analysis: Details
(2) Organize output. In the Output folder, MATS will generate a folder using the
Scenario Name. MATS will use this folder as a default location for files generated with
this Scenario Name.
r
C:\Program FileslAbt Associates\MATS\output
030
File
Edit View Favorites Tools Help
0
Back T Search
Folders
HS*
I Address Q C:\Program Files\Abt Associates\MATS\output
V
i3 go
Folders
X
Name
0 Q Program Files
- .^) Abt Associates
B Q MATS
Q configs
Ir^ data
l£) help
) maps
¦Q Example 03
O Tutorial 03
Q Tutorial Visibility
El Ql sampledata
© l_j work
¦A
£2) Tutorial 03
.^Example 03
Q Tutorial Visibility
BilTutorial 03.asr
]Safl Example 03.asr
[^Tutorial Visibility.asr
Si
Jj|.
¦
(3) Output file names. The output files generated will begin with the Scenario Name.
Abt Associates Inc.
306
June 2010
-------
Visibility Analysis: Details
\ Load j
Extract All
011®
Help T
Start Map View | Output Navigator I
Highlight file of interest and right-click to view options to Map, View, and Extract the data.
Name
i|Type
| Size
B Configuration/Log Files
j- Configuration
Configuration
54kb
s- Log File
Run Log
Okb
S Output Files
j- Tutorial Visibility-Class 1 Area and IMPROVE Monitor Identifiers and Locations
Monitor Network
17kb
; Tutorial Visibility - Forecasted Visibility - all design values
Monitor Network
6kb
f Tutorial Visibility - Forecasted Visibility Data
Monitor Network
2kb
j- Tutorial Visibility - Used Model Reference Cells - Base Data
Monitor Network
35kb
Tutorial Visibility-Used Model Reference Cells - Future Data
Monitor Network
35kb
Stop Info
11.1.2 Forecast Visibility at Class I Areas
MATS provides a forecast of visibility in Class I Areas. The approach used has the
following steps:
Identify best & worst visibility monitor days in Base Model Year. Use monitored
total extinction data from a user-specified year to identify the 20 percent best and 20
percent worst visibility days at each Class I area. At this stage MATS is using extinction
values (measured in inverse megameters). By the end of this series of calculations,
MATS will convert these extinction visibility measures to deciviews.
Note that you specify the particular year, the Base Model Year, from the available
monitoring data in the Filtering window. MATS labels the year of monitoring data as
the Base Model Year, because this particular year of monitor data matches the
baseline model data, specified in the Data Input window.
Average best & worst baseline model days. Using baseline speciated model data
(specified in the Data Input window), average the 20 percent best visibility days and
then average the 20 percent worst visibility days at each Class I areas (matched with the
ambient data). The model data comes into MATS as speciated values measured in ug/m3
. Average these speciated values. When done there will be two averages (one "best" the
other "worst") for each species and these averages will be ug/m3.
Abt Associates Inc.
307
June 2010
-------
Visibility Analysis: Details
Average best & worst forecast model days. Calculate these same averages for the
forecast model data (specified in the Data Input window). That is, identify the same 20
percent best and 20 percent worst visibility days and then average the speciated model
data (measured in ug/m3).* When done there will be two averages (one "best" the other
"worst") for each species and these averages will be in ug/m3.
Calculate RRFs. Use the speciated (best & worst) averages from the baseline and
forecast model data to calculate two RRFs for each species. That is, there will be one
RRF for the 20 percent best visibility days and another for the 20 percent worst visibility
days at each Class I area. The "best" RRF is simply the ratio of the baseline "best"
average (measured in ug/m3) to the control scenario "best" average (measured in ug/m3).
The "worst" RRF is calculated in an analogous way. An RRF is unitless and there are
two for each species.
Identify best & worst visibility days in other monitored years. Using monitored total
extinction, identify the 20 percent best visibility days and the 20 percent worst visibility
days from the other available years of monitoring data. The default in MATS is that
there should be at least three valid years and one of those years should be the base
modeling year (the base meteorological year). (Monitor validity is discussed further in
the Valid Visibility Monitors section.)
Note that the 20 percent best days will occur on a different set of days for each year;
similarly, the 20 percent worst days will occur on a different set of days for each year.
Multiply RRF with speciated monitor data from each year. Multiply the species-
specific "best" RRF (unitless) with the "best" daily speciated monitor values (measured
in ug/m3) from each of the available years. Do analogous calculations for the worst days.
When done, there will the original (baseline) monitor values and an analogous set of
forecast values (equal to the baseline times the RRF).
Note that the RRF is based on best/worst days identified from the Base Model Year.
This same "Base Model Year" RRF is used with all of the valid monitor years. For
example, if the Base Model Year were 2001, then the RRF developed from 2001
modeling data will be applied to all valid data in the five year ambient base period.
Convert ug/m3 values to daily extinction values and sum to get total extinction. For
each day in each valid monitor year (for both the baseline and forecast), use either the
New IMPROVE equation or the Old IMPROVE equation to convert ug/m3 values to get
daily total extinction (measured in inverse megameters). After this calculation there will
be a set of total extinction values for the best and worst visibility days in each valid year
for both the baseline and the forecast.
Convert extinction to deciviews. For each valid year in both the baseline and forecast,
convert the best & worst daily averages from extinction (inverse megameters) to
deciviews (unitless). The formula for this conversion is as follows: Deciviews = 10*ln
(extinction/10)
Average daily best and worst days. For each valid year, average the daily deciview
Abt Associates Inc.
308
June 2010
-------
Visibility Analysis: Details
values from the 20 percent best visibility days and calculate the same average for the 20
percent worst visibility days. There will be up to five "best" averages and "worst" total
visibility measures (measured in deciviews) for both the baseline and the forecast.
Calculate final average. Average the valid best/worst yearly visibility measures. When
done there will be one "best" value and one "worst" value, measured in deciviews
(unitless), for both the baseline and forecast.
* The future days are the same as the base year days. The identification of the 20 percent best and
worst is solely based on the base year ambient data.
11.1.2.1 Old IMPROVE Equation
The Old IMPROVE equation is as follows:
bext = 3*f(RH)*AMM_S04
+ 3*f(RH)*AMM_N03
+ 4*OMC
+ 10*EC
+ CRUSTAL
+ 0.6*CM
+ RAYLEIGH.
where:
bext = total extinction (measured in inverse megameters)
FRH = term to account for enhancement of light scattering due to hydroscopic growth of
suflate and nitrate (unitless)
AMM_S04 = ammonium sulfate (ug/m3)
AMM_N03 = ammonium nitrate (ug/m3)
OMC = organic carbon mass (ug/m3) (OC*1.4)
EC = elemental carbon (ug/m3)
CRUSTAL = fine soil (ug/m3)
CM = coarse particulate matter (ug/m3)
RAYLEIGH = Rayleigh scattering. Accounts for natural scattering of light by gases in the
Abt Associates Inc.
309
June 2010
-------
Visibility Analysis: Details
atmosphere. Assumed to equal 10 inverse megameters at all locations.
Example Calculation Old IMPROVE Equation
The first column "bext" presents the calculated value given the following data.
bext
ID
LAT
LONG
DATE
FRH
CRUSTAL
AMM N03
OMC
EC
CM
AMM S04
71.04
ACAD1
44.3771
-68.261
20000101
3.22
0.22
1.02
2.05
1.12
2.99
3.09
23.70
ACAD1
44.3771
-68.261
20000105
3.22
0.12
0.11
0.38
0.07
0.89
1.01
34.08
ACAD1
44.3771
-68.261
20000108
3.22
0.13
0.24
0.95
0.15
1.69
1.58
37.86
ACAD1
44.3771
-68.261
20000112
3.22
0.14
0.22
0.69
0.19
4.48
1.89
31.26
ACAD1
44.3771
-68.261
20000115
3.22
0.16
0.19
0.72
0.19
2.65
1.33
39.77
ACAD1
44.3771
-68.261
20000119
3.22
0.18
0.60
1.44
0.30
0.95
1.49
42.24
ACAD1
44.3771
-68.261
20000122
3.22
0.46
0.37
0.80
0.16
15.77
1.44
11.1.2.2 New IMPROVE Equation
The New IMPROVE Equation has a number of additional terms, in relation to the Old
IMPROVE equation. In particular, it takes into account the different effects of small and
large sulfate, nitrate, and organic carbon particles. A separate equation defining small and
large is given below.
bext = 2.2*fs(RH)*[SMALL_AMM_S04] + 4.8*f1(RH)*[LARGE_AMM_S04]
+ 2.4* fs(RH) * [ SM ALLAMMN03 ] + 5.1*f1(RH)*[LARGE_AMM_N03]
+ 2.8 * [ SM ALLOMC ] + 6.1*[LARGE_OMC]
+ 10*EC
+ CRUSTAL
+ 0.6*CM
+ S SRAYLEIGH
+ 1.7* fss(RH) * SE A S ALT
+ 0.33*NO2.
where:
bext = total extinction (measured in inverse megameters)
fs(RH) = term to account for enhancement of light scattering due to hydroscopic growth of
small ammonium nitrate and ammonium sulfate (unitless)
f[(RH) = term to account for enhancement of light scattering due to hydroscopic growth of
large ammonium nitrate and ammonium sulfate (unitless)
Abt Associates Inc.
310
June 2010
-------
Visibility Analysis: Details
SMALL_AMM_S04 = small ammonium sulfate (ug/m3)
LARGE AMM S04 = large ammonium sulfate (ug/m3)
SMALL_AMM_N03 = small ammonium nitrate (ug/m3)
LARGE_AMM_N03 = large ammonium nitrate (ug/m3)
SMALL OMC = small organic carbon mass (ug/m3) (OC*1.8)
LARGE OMC = large organic carbon mass (ug/m3) (OC*1.8)
EC = elemental carbon (ug/m3)
CRUSTAL = fine soil (ug/m3)
CM = coarse particulate matter (ug/m3)
SS RAYLEIGH = Site-specific Rayleigh scattering (inverse megameters)
fss(RH) = term to account for enhancement of light scattering due to hydroscopic growth
of sea salt (unitless)
SEA_SALT = Sea salt (ug/m3)
N02 = Nitrogen dioxide levels (parts per billion). This term is assumed to be zero.
The apportionment of the total concentration of sulfate compounds into the concentrations
of the small and large size fractions is accomplished using the following equations:
[Large Sulfate] = [Total Sulfate]/20 ug/m3 x [Total Sulfate], for [Total Sulfate] < 20
ug/m3
[Large Sulfate] = [Total Sulfate], for [Total Sulfate] > 20 ug/m3
[Small Sulfate] = [Total Sulfate] - [Large Sulfate]
The same equations are used to apportion total nitrate and total organic mass
concentrations into the small and large size fractions.
Example Calculation New IMPROVE Equation
The first column "bext" presents the calculated value given the following data.
Abt Associates Inc.
311
June 2010
-------
Visibility Analysis: Details
bext
ID
LAT
LONG
DATE
FRH
FSRH
FLRH
FSSRH
SS RAYLEIGH
71.52
ACAD1
44.3771
-68.261
20000101
3.22
3.82
2.75
3.91
12
24.51
ACAD1
44.3771
-68.261
20000105
3.22
3.82
2.75
3.91
12
34.45
ACAD1
44.3771
-68.261
20000108
3.22
3.82
2.75
3.91
12
38.10
ACAD1
44.3771
-68.261
20000112
3.22
3.82
2.75
3.91
12
35.45
ACAD1
44.3771
-68.261
20000115
3.22
3.82
2.75
3.91
12
40.22
ACAD1
44.3771
-68.261
20000119
3.22
3.82
2.75
3.91
12
43.92
ACAD1
44.3771
-68.261
20000122
3.22
3.82
2.75
3.91
12
SEA SALT
CRUSTAL
AMM N03
OMC
EC
PM10
CM
AMM S04
LARGE OMC
SMALL OMC
0
0.22
1.02
2.63
1.12
11.05
2.99
3.09
0.35
2.29
0
0.12
0.11
0.49
0.07
2.72
0.89
1.01
0.01
0.48
0
0.13
0.24
1.22
0.15
4.94
1.69
1.58
0.07
1.15
0
0.14
0.22
0.89
0.19
7.82
4.48
1.89
0.04
0.85
0.55116
0.16
0.19
0.93
0.19
5.53
2.65
1.33
0.04
0.89
0
0.18
0.60
1.85
0.30
5.03
0.95
1.49
0.17
1.68
0.192906
0.46
0.37
1.02
0.16
19.56
15.77
1.44
0.05
0.97
LARGE AMM S04
SMALL AMM S04
LARGE AMM N03
SMALL AMM N03
0.48
2.61
0.05
0.97
0.05
0.96
0.00
0.11
0.13
1.46
0.00
0.24
0.18
1.71
0.00
0.22
0.09
1.24
0.00
0.19
0.11
1.38
0.02
0.59
0.10
1.34
0.01
0.37
11.1.2.3 Choose Model Grid Cell
The model data are used to calculate an RRF for each species for both the best and worst
visibility days. The RRF is the ratio of future-year modeled visibility levels over baseline
modeled visibility levels. When forecasting visibility, MATS allows you to choose
whether the RRF will be calculated with model data from the grid cell containing the
monitor or the centroid of the Class I Area.
The representative IMPROVE monitor assignments are taken from Appendix A, Table A-2
of "Guidance for Tracking Progress Under the Regional Haze Rule"
http://www.epa.gov/ttn/oarpg/tl/memoranda/rh tpurhr gd.pdf. Monitors assigned to
represent a Class I Area are generally close to the Class I Area. However, in some cases,
this distance can be substantial. For example, the YELL2 monitor in Wyoming (44.5653
latitude, -110.4002 longitude) is located more than a degree longitude away from the Red
Rocks Lake Class I Area (44.64 latitude, -111.78 longitude). By default, MATS uses
model data at the monitor.
11.1.3 Visibility Output Variable Description
MATS generates five output files:
Abt Associates Inc.
312
June 2010
-------
Visibility Analysis: Details
Visibility forecast (average of design values). (Up to) five year average of forecasted
and base-year average visibility. When you have specified the option Use model grid
cells at monitor, name of this file is "Forecasted Visibility Data.csv" with the Scenario
Name appended at the beginning (e.g., "Tutorial Visibility - Forecasted Visibility Data,
csv"). When you have specified the option Use model grid cells at Class 1 area
centroid, name of this file changes to "Forecasted Visibility Data for Class 1 Areas, csv"
plus the Scenario Name (e.g., "Example Visibility - Forecasted Visibility Data for Class
1 Areas.csv").
Visibility forecast (all design values). Forecasted and base-year values for individual
years. (The forecast is based on The name of this file is "Forecasted Visibility - all
design values.csv" plus the Scenario Name (e.g., "Tutorial Visibility - Forecasted
Visibility - all design values.csv").
Class I areas and the monitors. Contains a list of all of the Class I areas and the monitors
assigned to each. The name of this file is "Class 1 Area and IMPROVE Monitor
Identifiers and Locations.csv" plus the Scenario Name (e.g., "Tutorial Visibility - Class 1
Area and IMPROVE Monitor Identifiers and Locations.csv").
Base-year model data used. The name of this file is: "Used Model Grid Cells - Base
Data.csv" plus the Scenario Name (e.g., "Tutorial Visibility - Used Model Grid Cells -
Base Data.csv"). This file contains the base year model values for PM species for the
grid cells and days used in the RRF calculations.
Future-year model data used. The format for this file is the same as for the base-year.
The name of this file is: "Used Model Grid Cells - Future Data.csv" plus the Scenario
Name (e.g., "Tutorial Visibility - Used Model Grid Cells - Future Data.csv"). This file
contains the future year model values for PM species for the grid cells and days used in
the RRF calculations.
The following sub-sections describe the variables in each file.
11.1.3.1 Forecasted Visibility Data.csv
The table below describes the variables in the output file. Note that the output data
includes a large number of variables, so in the sample output below we have divided the
variables into two blocks. In a file actually generated by MATS, these two blocks would
be combined.
Note also that the variables output by MATS depend on whether you have specified using
model data at the monitor or model data at the center of the Class I Area. This is detailed
in the description table below.
Variable Description
Jd Site ID
_type Leave blank
date Meteorological modeling year (used to identify the 20% best and worst
days from the ambient data)
Abt Associates Inc. June 2010
313
-------
Visibility Analysis: Details
dv_best
dv_worst
Forecasted (future year) best visibility [up to five year average] (in
deciviews)
Forecasted (future year) worst visibility [up to five year average]
base_best
Base-year best visibility [up to five year average]
base_worst
Base-year worst visibility [up to five year average]
rrf_b_crustal
rrf_b_no3
Relative response factor (RRF) for crustal matter on the best visibility
days
RRF for nitrate on the best visibility days
rrf_b_oc
RRF for organic mass on the best visibility days
rrf_b_ec
RRF for elemental carbon on the best visibility days
rrf_b_cm
RRF for coarse matter (PM10 minus PM2.5) on the best visibility days
rrf_b_so4
RRF for sulfate on the best visibility days
rrf_w_crustal
RRF for crustal matter on the worst visibility days
rrf_w_no3
RRF for nitrate on the worst visibility days
rrf_w_oc
RRF for organic mass on the worst visibility days
rrf_w_ec
RRF for elemental carbon on the worst visibility days
rrf_w_cm
RRF for coarse matter (PM10 minus PM2.5) on the worst visibility days
rrf_w_so4
RRF for sulfate on the worst visibility days
monitor_gridcell
Identifier for grid cell closest to monitor. (This variable only appears if you
specified the Use model grid cell at monitor option.)
cl ass_i_g rid ce 11
Identifier for grid cell closest to Class 1 area. (This variable only appears
if you specified the Use model grid cell at Class 1 area centroid option.)
gridcelljat
gridcelljong
monitorjat
Centroid latitude in decimal degrees of grid cell used in calculation. Values
in the northern hemisphere are positive, and those in the southern
hemisphere are negative.
Centroid longitude in decimal degrees of grid cell used in calculation.
Values in the eastern hemisphere are positive, and those in the western
hemisphere (e.g., United States) are negative.
Monitor latitude. (This variable only appears if you specified the Use
model grid cell at monitor option.)
class_i_lat
Class 1 area centroid latitude. (This variable only appears if you specified
the Use model grid cell at Class 1 area centroid option.)
monitorjong
Monitor longitude. (This variable only appears if you specified the Use
model grid cell at monitor option.)
class_i_long
Class 1 area centroid longitude. (This variable only appears if you
specified the Use model grid cell at Class 1 area centroid option.)
11.1.3.2 Forecasted Visibility - all design values.csv
The table below describes the variables in the output file. Note that the variables output by
MATS depend on whether you have specified using model data at the monitor or model
data at the center of the Class I Area. This is detailed in the description table below.
Variable Description
Jd Site ID
_type Leave blank
Abt Associates Inc.
314
June 2010
-------
Visibility Analysis: Details
date Base year of monitoring data
dv_best Forecasted (future year) best visibility (in deciviews)
dv_worst Forecasted (future year) worst visibility
base_best Base-year best visibility
base_worst Base-year worst visibility
monitor_gridceldentifier for grid cell closest to monitor. (This variable only appears if you specified
II the Use model grid cell at monitor option.)
class_i_gridcel Identifier for grid cell closest to Class 1 area. (This variable only appears if you
I specified the Use model grid cell at Class 1 area centroid option.)
gridcelljat Centroid latitude in decimal degrees of grid cell used in calculation. Values in the
northern hemisphere are positive, and those in the southern hemisphere are
negative.
gridcelljong Centroid longitude in decimal degrees of grid cell used in calculation. Values in the
eastern hemisphere are positive, and those in the western hemisphere (e.g., United
States) are negative.
monitorjat Monitor latitude. (This variable only appears if you specified the Use model grid
cell at monitor option.)
classjjat Class 1 area centroid latitude. (This variable only appears if you specified the Use
model grid cell at Class 1 area centroid option.)
monitorjong Monitor longitude. (This variable only appears if you specified the Use model grid
cell at monitor option.)
classjjong Class 1 area centroid longitude. (This variable only appears if you specified the Use
model grid cell at Class 1 area centroid option.)
11.1.3.3 Class 1 Area and IMPROVE Monitor Identifiers and Locations.csv
The table below describes the variables in the output file. Note that the output data
includes a number of variables, so in the sample output below we have divided the
variables into two blocks. In a file actually generated by MATS, these two blocks would
be combined.
Variable
Jd
Jype
_class_i_name
class_i_state
class_i_lat
class_i_long
class_i_g rid ce 11
date
_monitor_id
monitorjat
monitorjong
monitor_gridcell
Description
Class I area site ID
Leave blank.
Class 1 area name
State of Class 1 area
Latitude in decimal degrees of Class 1 area centroid
Longitude in decimal degrees of Class 1 area centroid
Identifier of grid cell closest to the Class 1 area centroid
Meteorological modeling year
IMPROVE site code (either at Class I area or a representative site)
IMPROVE Monitor latitude
IMPROVE Monitor longitude
Identifier of grid cell closest to the monitor
Abt Associates Inc.
315
June 2010
-------
Visibility Analysis: Details
11.1.3.4 Used Model Grid Cells - Base/Future Data.csv
The table below describes the variables in the output file.
Variable Description
Jd The ID is a unique name for each monitor in a particular location. The default
value is the column identifier multiplied by 1000 plus the row. (This is a character
variable.)
_type Leave blank
gridcelljat Latitude at the grid cell centroid in decimal degrees. Values in the northern
hemisphere are positive, and those in the southern hemisphere are negative.
gridcelljong Longitude at the grid cell centroid in decimal degrees. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g., United
States) are negative.
date Date of daily average model value with YYYYMMDD format (This is a numeric
variable)
cm Coarse PM (ug/m3)
crustal Crustal PM
so4 Sulfate PM
ec Elemental Carbon
no3 Nitrate PM
oc Organic carbon PM
_visibility_rank worst = 20% worst days used in rrf calculation; best = 20% best days used in rrf
calculation
11.2 Data Input
In the Data Input window, you need to specify the monitor data and model data that you
want to use. MATS comes with monitor and model data. Alternatively, you can add your
own, following the monitor and model format described below. In addition, you need to
specify how MATS will evaluate the model data when calculating RRFs.
Abt Associates Inc.
316
June 2010
-------
Visibility Analysis: Details
¦ Choose Desired Output
Data Input
¦ Filtering
ESI Final Check
Data Input
Monitor Data
IMPROVE Monitor Data • Old Algorithm JCAProgrann Files^Abt Associates\MATS\Sarnp ¦¦¦
IMPROVE Monitor Data ¦ New Algorithm |006-daily IMPROVE-all data-new equation.csv| ¦¦¦!
Model Data
Baseline File
Forecast File
Using Model Data
T emporal adjustment at monitor
|S\SampleData\2002cc_EUS_PM25_sub.csv
|S\SampleData\2020cc_ELIS_PM25_sub.csv
3x3 "r
< Back
Next >
Cancel
11.2.1 Monitor Data Input
Daily monitor data, with concentration (ug/m3) and visibility (inverse megameters)
measures for each species, are available from the VIEWS website
http://vista.cira.colostate.edu/views/.* As described in the Forecasting Visibility section,
these monitor data are used to: (1) identify the 20 percent best and worst visibility days for
a given year, and (2) calculate the 5-year baseline conditions.
Note that one IMPROVE monitor is associated with each Class I site, and the calculated
visibility for the IMPROVE site is assumed to representative of the Class I site. MATS
comes with a default "linkage" database that provides the cross-walk that MATS uses for
IMPROVE monitors and Class I Areas.
The tables in the next sub-sections present the old equation and new equation variable
names and descriptions downloaded from the VIEWS website and the variable names used
in MATS. The format the data read into MATS is also included.
By default, MATS includes species concentrations (measured in ug/m3) as well as
extinction estimates (measured in inverse megameters) and deciview values. MATS uses
the concentration estimates. In particular, it uses the variables: AMM_S04, AMM_N03,
OMC, EC, CRUSTAL, and CM. The variable GOOD YEAR indicates whether a
particular monitor should be used for a given year. A value of "1" means the monitor can
be used, and a value of "0" means that the monitor should be dropped for the year.
Abt Associates Inc.
317
June 2010
-------
Visibility Analysis: Details
The variable GROUP identifies the percentile for the overall visibility level for a particular
day. A value of "90" means that the particular day is among the 20 percent worst days for
the year. A value of "10" means that the particular day is among the 20 percent best days
for that year. (Days with other GROUP values are not needed)
There are a number of extra variables in the ambient data input file that are not directly
used by MATS (such as extinction values). The additional data can be used to QA MATS
output or for additional data analysis.
* The Visibility Information Exchange Web System (VIEWS) is an online exchange of air quality data,
research, and ideas designed to understand the effects of air pollution on visibility and to support the
Regional Haze Rule enacted by the U.S. Environmental Protection Agency (EPA) to reduce regional
haze and improve visibility in national parks and wilderness areas, http://vista.cira.colostate.edu/views/
11.2.1.1 Monitor Data Description (Old Equation)
The monitor data for the old IMPROVE algorithm includes a large number of variables, so
in the sample format below we have only a portion of the variables listed. The table below
has a complete listing of the variables. The variables in bold are required variables on the
input file. The additional variables are provided for QA purposes and are optional.
Visibility Monitor Data Format (Old Algorithm)
|°ay
_ID, _TYPE, LAT, LONG, DATE, FRH, PM25, CRU5TAL, AMM_N03, OMC, EC, PM10, CM, AM
"ACAD1",44.3771,-68.261,20000101,3.22,8.0645,0.2171958,1.017423,2.04764,1.11
"ACADl","", 44.3771,-68.261,20000105,3 .22,1.8300, 0 .1202492, 0 .111198, 0 .3829,0.067
"ACADl","",44.3771,-68.261,2 0000108,3.22,3.2492,0.1289628,0.240972,0.95102,0.14
"ACAD1","",44.3771,-68.261,2 0000112,3.22,3.3448,0.144354,0.2193,0.693 8 4,0.18 66,
"ACADl","",44.37 71,-68.261,2 0000115,3.22,2.8856,0.1553525,0.187308,0.72184,0.19
"ACADl","",44.3771,-68.261,20000119,3.22,4.0888,0.1827762,0.604623,1.44088,0.30
"ACADl","", 44.3771,-68.261,20000122,3 .22,3.7937, 0 .4609729, 0 .372681, 0 .79506, 0 .16
"ACADl","",44.3771,-68.261,20000126,3.22,7.9274,0.1164544,0.807927,1.19252,0.14 v
< JIN I >
Visibility Monitor Data Variables and Descriptions (Old Algorithm) [Variables in
bold are required]
Variable Description
_ID IMPROVE site code
_TYPE Leave blank
LAT Latitude in decimal degrees. Values in the northern hemisphere are positive,
and those in the southern hemisphere are negative.
Abt Associates Inc. June 2010
318
-------
Visibility Analysis: Details
LONG
Longitude in decimal degrees. Values in the eastern hemisphere are positive,
and those in the western hemisphere (e.g., United States) are negative.
DATE
Date of daily average ambient data with YYYYMMDD format (This is a
numeric variable)
FRH
Monthly climatological relative humidity adjustment factor
PM25
Measured PM2.5 mass (ug/m3)
CRUSTAL
Crustal mass (2.2 x [Al] + 2.49 x [Si] + 1.63 x [Ca] + 2.42 x [Fe] + 1.94 x [Ti])
AMM_N03
Ammonium nitrate mass (N03*1.29)
OMC
Organic carbon mass (OC*1.4)
EC
Elemental carbon
PM10
PM10 mass
CM
Coarse mass (PM10 minus PM2.5)
AMM_S04
Ammonium sulfate (S*4.125)
E_AMM_S04
Ammonimum sulfate extinction (Mm-1)
E_AMM_N03
Ammonimum nitrate extinction
E_OMC
Organic mass extinction
E_EC
Elemental carbon extinction
E_CRUSTAL
Crustal extinction
E_CM
Coarse mass (PM10 minus PM2.5) extinction
TBEXT
Total bext (includes 10 Mm-1 for Rayleigh scattering)
DV
Deciviews (calculated from Total bext)
GOOD_YEAR
Denotes complete data for the year (1= all quarters >75% completeness, 0=
incomplete)
GROUP
90= 20% worst days and 10= 20% best days for each year (if good_year= 1)
POSSIBLE_NDAYSPossible samples in quarter
NDAYS Actual complete samples per quarter
Quarter completeness (1= complete, 0= incomplete)
COMPLETE_QUAR
TER
SF Sulfur concentration (used to calculate ammonium sulfate)
S04F Sulfate concentration (may be used as a backup in case S is missing)
Note: Character variables have names that begin with an underscore {i.e., and the
character values used can be kept with or without quotes. (If a character variable has an
embedded space, such as might occur with the name of a location, then use quotes.)
11.2.1.2 Monitor Data Description (New Equation)
The monitor data for the new IMPROVE algorithm includes a large number of variables,
so in the sample format below we have only a portion of the variables listed. The table
below has a complete listing of the variables. The variables in bold are required variables
on the input file. The additional variables are provided for QA purposes and are optional.
Note that sea salt mass is optional, but is used in the visibility calculations if provided.
Visibility Monitor Data Format (New Algorithm)
Abt Associates Inc.
319
June 2010
-------
Visibility Analysis: Details
Day
_ID,_TYPE,LAT,LONG,DATE,FRH,F5RH,FLRH,FSSRH,SS_RAYLEIGH,SEA_SALT,PM25,CRUSTAL,A
ACAD1,,44.3771,-68 .261,20000101,3 .22,3 .82,2.75,3 .91,12, 0, 8 .0645,0.2171958,1.017
ACAD1,,44.3771,-68.261,20000105,3.22,3.82,2.75,3.91,12,0,1.8308,0.1202492,0.111
ACAD1,,44.3771,-68.261,20000108,3.22,3.82,2.75,3.91,12,0,3.2492,0.1289628,0.240
ACAD1,,44.37 71,-68.261,20000112,3.22,3.82,2.75,3.91,12,0,3.3448,0.144354,0.2193
ACAD1,,44.3771,-68.261,20000115,3.22,3.82,2.75,3.91,12,0.55116,2.8856,0.1553525
ACAD1,, 44.3771,-68 .261,20000119,3.22,3 .82,2.75,3 .91,12, 0, 4.0888,0.1827762,0 .604
ACAD1,,44.3771,-68.261,20000122,3.22,3.82,2.75,3.91,12,0.192906,3.7937,0.460972
ACAD1,,44.37 71,-68.261,20000126,3.22,3.82,2.75,3.91,12,2.29752,7.9274,0.1164544 v|
Visibility Monitor Data Variables and Descriptions (New Algorithm) [Variables in
bold are required]
Variable
ID
TYPE
LAT
LONG
DATE
FRH
FSRH
FLRH
FSSRH
SS_RAYLEIGH
PM25
SEA_SALT
CRUSTAL
AMM_N03
OMC
EC
PM10
CM
AMM_S04
LARGE_OMC
SMALL_OMC
LARGE_AMM_S04
SMALL_AMM_S04
LARGE AMM NQ3
Description
IMPROVE site code
Leave blank
Latitude in decimal degrees. Values in the northern hemisphere are positive,
and those in the southern hemisphere are negative.
Longitude in decimal degrees. Values in the eastern hemisphere are positive,
and those in the western hemisphere (e.g., United States) are negative.
Date of daily average ambient data with YYYYMMDD format (This is a
numeric variable)
Monthly climatological relative humidity adjustment factor
Monthly climatological relative humidity adjustment factor - small sulfate and
nitrate particles
Monthly climatological relative humidity adjustment factor - large sulfate and
nitrate particles
Monthly climatological relative humidity adjustment factor - sea salt
Site-specific Rayleigh scattering (Mm-1)
Measured PM2.5 mass (ug/m3)
Sea salt mass
Crustal mass (2.2 x [Al] + 2.49 x [Si] + 1.63 x [Ca] + 2.42 x [Fe] + 1.94 x [Ti])
Ammonium nitrate mass (N03*1.29)
Organic carbon mass (OC*1.8)
Elemental carbon
PM10 mass
Coarse mass (PM10 minus PM2.5)
Ammonium sulfate (S*4.125)
Large organic mass
Small organic mass
Large ammonium sulfate
Small ammonium sulfate
Large ammonium nitrate
Abt Associates Inc.
320
June 2010
-------
Visibility Analysis: Details
SMALL_AMM_N03
E_AMM_S04
E_AMM_N03
E_OMC
E_EC
E_CRUSTAL
E_CM
E_SEA_SALT
TBEXT
DV
GOOD_YEAR
GROUP
POSSIBLE_NDAYS
NDAYS
COMPLETE_QUAR
TER
SF
S04F
Small ammonium nitrate
Ammonimum sulfate extinction (Mm-1)
Ammonimum nitrate extinction
Organic mass extinction
Elemental carbon extinction
Crustal extinction
Coarse mass (PM10 minus PM2.5) extinction
Sea salt extinction
Total bext (includes site specific Rayleigh scattering)
Deciviews (calculated from Total bext)
Denotes complete data for the year (1= all quarters >75% completeness, 0=
incomplete)
90= 20% worst days and 10= 20% best days for each year (if good_year= 1)
Possible samples in quarter
Actual complete samples per quarter
Quarter completeness (1= complete, 0= incomplete)
Sulfur concentration (used to calculate ammonium sulfate)
Sulfate concentration (may be used as a backup in case S is missing)
Note: Character variables have names that begin with an underscore (i.e., and the
character values used can be kept with or without quotes. (If a character variable has an
embedded space, such as might occur with the name of a location, then use quotes.)
11.2.1.3 Linkage between Monitors & Class I Areas
MATS comes with a default database that provides the linkage between IMPROVE
monitors and Class I Areas. The file is called "156-Class I-coordinates-all site names.csv".
The format of the file and the variable descriptions are as follows:
Format for File Linking IMPROVE Monitors and Class I Areas
Abt Associates Inc.
321
June 2010
-------
Visibility Analysis: Details
Variables and Descriptions for File Linking IMPROVE Monitors and Class I Areas
Var Name
Description (variable type)
MONITOR ID
IMPROVE monitor identification code (text)
MonLAT
IMPROVE monitor latitude (numeric)
MonLONG
IMPROVE monitor longitude (numeric)
_CL AS SINAME
Name of Class I Area (text)
ID
Class I Area identification code (text)
STATEID
State in which Class I Area is located (text)
LAT
Class I Area centroid latitude (numeric)
LONG
Class I Area centroid longitude (numeric)
Note: Character variables have names that begin with an underscore {i.e., and the
character values used can be kept with or without quotes. (If a character variable has an
embedded space, such as might occur with the name of a location, then use quotes.)
11.2.2 Model Data Input
The model data for the 20 percent best and worst visibility days are used to calculate
relative response factors (RRFs), which provide an estimate of the relative change in
visibility from the baseline conditions to a future year. Recall from forecast steps that the
monitor data are used to identify the best and worst days. Not the model data. MATS will
match the best and worst measured days to the correct modeled days, by date. The model
data input to MATS is in terms of PM species concentrations (measured in ug/m3).
The following exhibits provide an example of the model data format and a description of
the variables. Note that the first line of the data file gives the frequency of the data. In this
case, daily data. The second line gives the variables names. The data begins on the third
line. Each data line represents a daily observation.
Format for Daily PM Model Data
Abt Associates Inc.
322
June 2010
-------
Visibility Analysis: Details
0
Day
A
o
JED,
_TYPE, LAT, LONG, DATE, CH, CRUSTAL, S04, EC, N03, 0C
o
1001
"",18.362337,
-121.659843,20150101,0.0001,0.0003,0.0448,0.0037,0,0.0168
0
1001
"",18.362337,
-121.659843,20150102,0.0058,0.0166,0.1202,0.0072,0,0.0325
o
1001
"",18.362337,
-121.659843,20150103,0.0001,0.0001,0.0838,0.0093,0,0.0412
o
1001
"",18.362337,
-121.659843,20150104,0,0,0.2813,0.0118,0,0.0481
o
1001
"",18.362337,
-121.659843,20150105,0,0,0.6533,0.0354,0.0008,0.1588
o
1001
"",18.362337,
-121.659843,20150106,0,0,0.5854,0.0249,0.0005,0.1146
o
1001
"",18.362337,
-121.659843,20150107,0,0.0001,0.8663,0.0228,0.0003,0.1017
o
1001
"",18.362337,
-121.659843,20150108,0.0011,0.0056,0.4761,0.0147,0,0.0658
o
1001
"",18.362337,
-121.659843,20150109,0.0153,0.057,0.3192,0.0157,0,0.0751
0
1001
"",18.362337,
-121.659843,20150110,0.005,0.0217,0.1882,0.0123,0,0.048
o
1001
"",18.362337,
-121.659843,20150111,0,0,0.0995,0.007,0,0.0219
o
1001
"",18.362337,
-121.659843,20150112,0.0018,0.0022,0.1087,0.0072,0,0.0235
0
1001
"",18.362337,
-121.659843,20150113,0.0015,0.0022,0.1335,0.008,0,0.027
o
1001
"",18.362337,
-121.659843,20150114,0.0002,0.0006,0.1797,0.008,0,0.0256
o
1001
"",18.362337,
-121.659843,20150115,0.0025,0.0061,0.1131,0.0069,0,0.0267
o
1001
"",18.362337,
-121.659843,20150116,0.0189,0.0407,0.2609,0.016,0,0.0658
o
1001
"",18.362337,
-121.659843,20150117,0.0011,0.0021,0.2342,0.0157,0,0.0532
V
<
>
Visibility Model Data Variable Descriptions
Variable Description
_ID The ID is a unique number for each model grid cell in the air quality model domain.
It is generally based on the column and row identifiers from the air quality modeling
domain. The default convention is to calculate the ID by multiplying the column
identifier by one thousand (1000) and adding the row identifier. (This is a character
variable.)
_TYPE Leave blank
LAT Latitude in decimal degrees of the center of each grid cell. Values in the northern
hemisphere are positive, and those in the southern hemisphere are negative.
LONG Longitude in decimal degrees of the center of each grid cell. Values in the eastern
hemisphere are positive, and those in the western hemisphere (e.g., United States)
are negative.
DATE Date of daily average model value with YYYYMMDD format (This is a numeric
variable)
CM Coarse PM (ug/m3)
CRUSTAL Crustal PM
S04 Sulfate PM
EC Elemental carbon
N03 Nitrate PM
OC Organic mass PM
Note: Character variables have names that begin with an underscore (i.e., and the
character values used can be kept with or without quotes. (If a character variable has an
embedded space, such as might occur with the name of a location, then use quotes.)
Finally, note that the species variables used by MATS do not exactly correspond to the
Abt Associates Inc.
323
June 2010
-------
Visibility Analysis: Details
speciated monitor data input available from the VIEWS website. The following exhibit
presents the correspondence used by MATS.
Monitor Data
Variable Name (from Views) New Variable Name (used within
MATS)
Soil
CRUSTAL
E
E
<
_N03
AMM_N03
OMC
OMC
LAC
EC
CM
CM
E
E
<
_S04
AMM_S04
Regardless of the species names used by the air quality model, the model output variables
should be changed to the MATS variable names when creating MATS input files.
11.2.2.1 Using Model Data for Temporal Adjustment
Relative response factors (RRFs) are calculated for each species: sulfate, nitrate, EC,
OMC, Crustal, and Coarse Matter (CM), by taking the ratio of the average of the 20
percent best (or worst) days in the future to the average of the 20 percent best (or worst)
days in the baseline. For example, when calculating the sulfate RRF for the 20 percent
best days, MATS does the following calculation:
1 ™
= ~j n
-¦ J#
n !-l
where:
j = Class I area
i = day identifier
n = number of 20 percent best visibility days
Sulfate = modeled sulfate concentration (in ug/m3) on best visibility days.
When identifying the model data for this calculation, MATS first determines whether you
want to use model values located at the monitor or at the centroid of the Class I Area. You
choose the desired location (monitor or cell centroid) in the Choose Desired Output
window. In addition, you need to specify how many cells around the desired location you
Abt Associates Inc.
324
June 2010
-------
Visibility Analysis: Details
want to use in the calculation (lxl matrix, 3x3 matrix, etc), and whether you want to use
the maximum or the mean of the model cells.
0
Data Input
Monitor Data
IMPROVE Monitor Data- Old Algorithm |TS\SarnpleData\visibility_monitor_data.csv
-]
IMPROVE Monitor Data-New Algorithm 04-daily IMPROVE-all data-new equation.csv 3D
Model Data
Baseline File
Forecast File
Using Model Data
Temporal adjustment at monitor
|ATS\SampleData\visibility_model_gQCI1 .csv •""]
|ATS\SampleData\visibility_model_2015.csv 3D
1x1 Mean :
3x3
5x5
7x7
< Back
Next >
Cancel
In the case of a 3x3 matrix with Mean specified, MATS identifies the speciated model
values (measured in ug/m3) from among the nine "nearby" grid cells for each day for each
Class I Area. In the typical case, where there are 365 days of model outputs, MATS will
generate 365 daily values. MATS will do this calculation separately for each species for
both the baseline and future-year scenarios. The Guidance Document recommends using
the Mean of model values when calculating the RRF. Next there is a recommended
example of how MATS calculates the RRF using the Mean for a 3x3 matrix. (An example
with the Maximum is also provided, however this is not the recommended approach for
visibility calculations.)
Abt Associates Inc.
325
June 2010
-------
Visibility Analysis: Details
RRF Calculation - Example with Mean
The Guidance Document recommends using the Mean of the model values. In the case of
a 3x3 matrix with Mean specified, MATS averages the speciated model values (measured
in ug/m3) from the nine "nearby" grid cells for each day for each Class I Area.
Assume there are eight best visibility days with the following modeled sulfate values in
the baseline. MATS would average the values for each day.
Best Days
Cell 1
Cell 2
Cell 3
Cell 4
Cell 5
Cell 6
Cell 7
Cell 8
Cell 9
Mean
1
1.795
1.812
1.299
1.609
1.612
1.250
1.692
0.570
1.347
1.443
2
0.164
1.556
1.205
0.270
1.940
1.065
1.156
1.620
1.786
1.196
3
1.709
0.273
1.304
1.213
1.177
1.104
0.368
1.817
1.377
1.149
4
1.119
1.322
1.778
1.154
1.503
1.511
1.251
1.939
0.474
1.339
5
1.910
1.648
1.012
1.635
1.912
1.587
1.508
1.723
1.611
1.616
6
1.490
1.204
1.997
0.989
1.832
0.064
1.469
1.634
1.470
1.350
7
1.136
1.886
1.131
1.282
1.957
1.047
1.335
0.045
1.279
1.233
8
1.304
1.217
1.738
1.243
1.370
1.802
1.374
1.736
1.196
1.442
Assume there are eight best visibility days with the following modeled sulfate values in
the forecast. Again, MATS would average the values for each day.
Best Days Cell 1 Cell 2 Cell 3 Cell 4 Cell 5 Cell 6 Cell 7 Cell 8 Cell 9 Max
1 1.789 1.715 1.209 1.560 1.562 1.224 1.492 0.489 1.148 1.354
2 0.137 1.512 1.162 0.181 1.939 1.022 1.113 1.541 1.593 1.133
Abt Associates Inc.
326
June 2010
-------
Visibility Analysis: Details
3
1.695
0.208
1.254
1.198
1.133
1.102
0.200
1.244
1.235
1.030
4
1.090
1.251
1.627
1.126
1.470
1.468
1.120
1.877
0.325
1.262
5
1.815
1.552
0.974
1.549
1.594
1.546
1.407
1.707
1.591
1.526
6
1.327
1.167
1.880
0.957
1.756
0.000
1.318
1.590
1.392
1.265
7
0.989
1.805
1.028
1.212
1.820
1.010
1.183
0.042
1.236
1.147
8
1.127
1.183
1.673
1.238
1.291
1.753
1.220
1.717
1.004
1.356
The average across the daily means for the baseline is 1.346 ug/m3. The average of the
forecast cells is 1.259. The sulfate RRF would then be calculated as: RRF = 1.259 / 1.346
= 0.935.
A similar process occurs for the other species. The end result is 12 RRFs — two for each
of the six species {i.e., sulfate, nitrate, elemental carbon, organic carbon, crustal, and
ammonium).
RRF Calculation - Example with Maximum
The Modeling Guidance Document recommends using the Mean of the model values for
visibility calculations. The example below, shows the calculations that would be involved
if the Maximum were chosen. In the case of a 3x3 matrix with Maximum specified, MATS
identifies identify the highest speciated model values (measured in ug/m3) from among the
nine "nearby" grid cells for each day for each Class I Area.
Assume there are eight best visibility days with the following modeled sulfate values in
the baseline:
Best Days
Cell 1
Cell 2
Cell 3
Cell 4
Cell 5
Cell 6
Cell 7
Cell 8
Cell 9
1
1.795
1.812
1.299
1.609
1.612
1.250
1.692
0.570
1.347
2
0.164
1.556
1.205
0.270
1.940
1.065
1.156
1.620
1.786
3
1.709
0.273
1.304
1.213
1.177
1.104
0.368
1.817
1.377
4
1.119
1.322
1.778
1.154
1.503
1.511
1.251
1.939
0.474
5
1.910
1.648
1.012
1.635
1.912
1.587
1.508
1.723
1.611
6
1.490
1.204
1.997
0.989
1.832
0.064
1.469
1.634
1.470
7
1.136
1.886
1.131
1.282
1.957
1.047
1.335
0.045
1.279
8
1.304
1.217
1.738
1.243
1.370
1.802
1.374
1.736
1.196
MATS would choose the cells highlighted in orange:
Best Days
Cell 1
Cell 2
Cell 3
Cell 4
Cell 5
Cell 6
Cell 7
Cell 8
Cell 9
1
1.795
1.812
1.299
1.609
1.612
1.250
1.692
0.570
1.347
2
0.164
1.556
1.205
0.270
1.940
1.065
1.156
1.620
1.786
3
1.709
0.273
1.304
1.213
1.177
1.104
0.368
1.817
1.377
4
1.119
1.322
1.778
1.154
1.503
1.511
1.251
1.939
0.474
5
1.910
1.648
1.012
1.635
1.912
1.587
1.508
1.723
1.611
6
1.490
1.204
1.997
0.989
1.832
0.064
1.469
1.634
1.470
7
1.136
1.886
1.131
1.282
1.957
1.047
1.335
0.045
1.279
8
1.304
1.217
1.738
1.243
1.370
1.802
1.374
1.736
1.196
Abt Associates Inc.
327
June 2010
-------
Visibility Analysis: Details
Assume there are eight best visibility days with the following modeled sulfate values in
the forecast:
I Days
Cell 1
Cell 2
Cell 3
Cell 4
Cell 5
Cell 6
Cell 7
Cell 8
Cell 9
1
1.789
1.715
1.209
1.560
1.562
1.224
1.492
0.489
1.148
2
0.137
1.512
1.162
0.181
1.939
1.022
1.113
1.541
1.593
3
1.695
0.208
1.254
1.198
1.133
1.102
0.200
1.744
1.235
4
1.090
1.251
1.627
1.126
1.470
1.468
1.120
1.877
0.325
5
1.815
1.552
0.974
1.549
1.894
1.546
1.407
1.707
1.591
6
1.327
1.167
1.880
0.957
1.756
0.000
1.318
1.590
1.392
7
0.989
1.805
1.028
1.212
1.820
1.010
1.183
0.042
1.236
8
1.127
1.183
1.673
1.238
1.291
1.753
1.220
1.717
1.004
MATS would choose the cells highlighted in orange. Note that the cells chosen for the
forecast can differ from the cells chosen for the baseline.
t Days
Cell 1
Cell 2
Cell 3
Cell 4
Cell 5
Cell 6
Cell 7
Cell 8
Cell 9
1
1.789
1.715
1.209
1.560
1.562
1.224
1.492
0.489
1.148
2
0.137
1.512
1.162
0.181
1.939
1.022
1.113
1.541
1.593
3
1.695
0.208
1.254
1.198
1.133
1.102
0.200
1.244
1.235
4
1.090
1.251
1.627
1.126
1.470
1.468
1.120
1.877
0.325
5
1.815
1.552
0.974
1.549
1.594
1.546
1.407
1.707
1.591
6
1.327
1.167
1.880
0.957
1.756
0.000
1.318
1.590
1.392
7
0.989
1.805
1.028
1.212
1.820
1.010
1.183
0.042
1.236
8
1.127
1.183
1.673
1.238
1.291
1.753
1.220
1.717
1.004
The average of the best sulfate days chosen from the baseline cells is 1.897 ug/m3. The
average of the forecast cells is 1.821. The sulfate RRF would then be calculated as: RRF
= 1.821 / 1.897 = 0.960.
A similar (independent) process occurs for the other species. The particular cells chosen
for sulfate may be quite different from the cells chosen for, say, nitrate. Continuing with
the example you might have the following pattern for baseline:
t Days
Cell 1
Cell 2
Cell 3
Cell 4
Cell 5
Cell 6
Cell 7
Cell 8
Cell 9
1
0.071
0.133
0.620
0.788
0.808
0.964
0.763
0.938
0.158
2
0.561
0.891
0.105
0.689
0.695
0.302
0.523
0.209
0.485
3
0.005
0.384
0.604
0.077
0.545
0.046
0.177
0.664
0.821
4
0.926
0.651
0.111
0.334
0.887
0.548
0.447
0.547
0.730
5
0.630
0.995
0.769
0.888
0.379
0.121
0.779
0.130
0.558
6
0.700
0.761
0.993
0.556
0.659
0.877
0.761
0.474
0.821
7
0.074
0.189
0.619
0.987
0.279
0.757
0.470
0.189
0.701
8
0.848
0.100
0.964
0.535
0.566
0.315
0.440
0.011
0.852
And the following pattern for the forecast:
Abt Associates Inc.
328
June 2010
-------
Visibility Analysis: Details
Best Days
Cell 1
Cell 2
Cell 3
Cell 4
Cell 5
Cell 6
Cell 7
Cell 8
Cell 9
1
0.068
0.122
0.556
0.711
0.766
0.594
0.653
0.877
0.152
2
0.514
0.515
0.092
0.615
0.687
0.283
0.460
0.198
0.479
3
0.004
0.380
0.513
0.076
0.504
0.044
0.175
0.591
0.801
4
0.769
0.581
0.099
0.327
0.852
0.537
0.416
0.482
0.719
5
0.563
0.940
0.673
0.859
0.374
0.113
0.685
0.123
0.505
6
0.610
0.694
0.910
0.514
0.621
0.828
0.649
0.438
0.794
7
0.072
0.179
0.597
0.954
0.244
0.701
0.428
0.176
0.695
8
0.837
0.090
0.940
0.522
0.500
0.295
0.434
0.011
0.809
The end result is 12 RRFs — two for each of the six species {i.e., sulfate, nitrate, elemental
carbon, organic carbon, crustal, and ammonium).
11.3 Filtering
MATS loads in the monitor data that you have specified in Data Input window, and then in
the Filtering window MATS presents the available years of monitor data for your analysis.
You specify a range of years with the Start Monitor Year and End Monitor Year
drop-down menus.
Using the Base Model Year drop-down menu, you can also specify the year that you want
to use to determine the "best" and "worst" monitor days. The Base Model Year needs to
fall within the range specified by the Start Monitor Year and End Monitor Year. Once
you have specified the Base Model Year, MATS will then identify and save for each
monitor the particular dates during this year that registered the best and worst visibility
days. These dates are then used to identify the model values used in the calculation of
RRFs for the temporal adjustment, as seen in the Example in the Model Data Input section.
Given a particular range of years that you have chosen, you can also specify the criteria for
a monitor to be included in the analysis. With the Minimum years required for a valid
monitor box, you specify the minimum number of years of data that a monitor must have (
e.g., three years).
You can also specify the Maximum Distance from Domain. That is, you can choose the
maximum distance that a monitor (or Class I Area centroid) can be from the nearest model
grid cell centroid. For example, if the Maximum Distance from Domain is 25, and a
monitor is more than 25 kilometers from the nearest model grid cell centroid, then a
forecast is not generated for this particular monitor. (More detailed examples regarding
distance and monitor validity are available here.)
Abt Associates Inc.
329
June 2010
-------
Visibility Analysis: Details
¦ Choose Desired Output
¦ Data Input
Filtering
ESI Final Check
Filtering
Choose Visibility Data Years
Start Monitor Year End Monitor Year Base Model Year
ISNHflQ 12004 T] 12002 V]
Valid Visibility Monitors
Minimum years required for a valid monitor 3
Man Distance from Centroid to Gridcell Center [km] 25
< Back
Next >
Cancel
11.3.1 Example Valid Visibility Monitors
Using the Maximum Distance from Domain, you can choose the maximum distance that
a monitor (or Class I Area centroid) can be from the nearest model grid cell centroid.
Whether you calculate the distance from a monitor or a Class I Area centroid depends on
whether you have specified Use model grid cells at monitor or Use model grid cells at
Class I area centroid.
Abt Associates Inc.
330
June 2010
-------
Visibility Analysis: Details
Point Estimates
Scenario Name : |
Forecast
w Tempo rally-adjust visibility levels at Class 1 Areas
IMPROVE Algorithm
(•* use old version C use new version
0 Use model grid cells at monitor
Use model grid cells at Class 1 area centroid
Example 1
Assume you have chosen Use model grid cells at monitor. If you have set the Maximum
Distance from Domain to 25, and a monitor is more than 25 kilometers from the nearest
model grid cell centroid, then a forecast is not generated for this particular monitor. And
by extension, a forecast is not generated for the Class I Areas that are associated with this
particular monitor.
Recall from the section on the Linkage between Monitors and Class I Areas that more than
one Class I Area may be linked to a monitor. Highlighted in yellow below are some
examples of monitors associated with more than Class I Area. For example, if the CHIR1
monitor is more than 25 kilometers from the nearest model grid cell centroid, then no
forecasts would be generated for the three Class I Areas associated with this monitor (i.e.,
Chiricahua NM, Chiricahua Wilderness, and Galiuro Wilderness).
MONITOR
MonLAT MonLONG_CLASS_l_NAME
ID
STATE
LAT LONG
ID
ID
ACAD1
44.377
-68.261
Acadia NP
ACAD
ME
44.35 -68.24
AGTI1
33.464
-116.971
Agua Tibia Wilderness
AGTI
CA
33.42 -116.99
BADL1
43.744
-101.941
Badlands NP
BADL
SD
43.81 -102.36
BALD1
34.058
-109.441
Mount Baldy Wilderness
BALD
AZ
33.95 -109.54
BAND1
35.780
-106.266
Bandelier NM
BAND
NM
35.79 -106.34
BIBE1
29.303
-103.178
Big Bend NP
BIBE
TX
29.33 -103.31
Abt Associates Inc.
331
June 2010
-------
Visibility Analysis: Details
BLIS1
38.976
-120.103
Desolation Wilderness
DESO
CA
38.9 -120.17
BLIS1
38.976
-120.103
Mokelumne Wilderness
MOKE
CA
38.57 -120.06
BOAP1
33.870
-106.852 Bosque del Apache
BOAP
NM
33.79 -106.85
BOWA1
47.947
-91.496
Boundary Waters Canoe Area
BOWA
MN
48.06 -91.43
BRCA1
37.618
-112.174
Bryce Canyon NP
BRCA
UT
37.57 -112.17
BRET1
29.119
-89.207
Breton
BRET
LA
29.87 -88.82
BRID1
42.975
-109.758
Bridger Wilderness
BRID
WY
42.99 -109.49
BRID1
42.975
-109.758
Fitzpatrick Wilderness
FITZ
WY
43.24 -109.6
BRIG1
39.465
-74.449
Brigantine
BRIG
NJ
39.49 -74.39
CABI1
47.955
-115.671
Cabinet Mountains Wilderness
CABI
MT
48.18 -115.68
CACR1
34.454
-94.143
Caney Creek Wilderness
CACR
AR
34.41 -94.08
CANY1
38.459
-109.821
Arches NP
ARCH
UT
38.73 -109.58
CANY1
38.459
-109.821
Canyonlands NP
CANY
UT
38.23 -109.91
CAPI1
38.302
-111.293 Capitol Reef NP
CAP I
UT
38.06 -111.15
CHAS1
28.748
-82.555
Chassahowitzka
CHAS
FL
28.69 -82.66
CHIR1
32.009
-109.389
Chiricahua NM
CHIR
AZ
32.01 -109.34
CHIR1
32.009
-109.389
Chiricahua Wilderness
CHIW
AZ
31.86 -109.28
CHIR1
32.009
-109.389
Galiuro Wilderness
GALI
AZ
32.6 -110.39
COHU1
34.785
-84.627
Cohutta Wilderness
COHU
GA
34.93 -84.57
CRLA1
42.896
-122.136
Crater Lake NP
CRLA
OR
42.92 -122.13
CRLA1
42.896
-122.136
Diamond Peak Wilderness
DIPE
OR
43.53 -122.1
CRLA1
42.896
-122.136
Gearhart Mountain Wilderness
GEMO
OR
42.51 -120.86
CRLA1
42.896
-122.136
Mountain Lakes Wilderness
MOLA
OR
42.33 -122.11
CRM01
43.461
-113.555
Craters of the Moon NM
CRMO
ID
43.39 -113.54
Example 2
Assume you have chosen Use model grid cells at Class I area centroid. If you have set
the Maximum Distance from Domain to 25, and a Class I Area is more than 25
kilometers from the nearest model grid cell centroid, then a forecast is not generated for
this particular Class I Area (e.g., Chiricahua, NM).
As noted above, the CHIR1 monitor is linked to two other Class I Areas. If these two other
areas are within 25 kilometers of a model grid cell centroid, then the monitor values from
CHIR1 would be used in the forecast for these two areas (along with the model values
associated with the centroid of each area).
11.4 Final Check
The Final Check window verifies the selections that you have made.
Abt Associates Inc.
332
June 2010
-------
Visibility Analysis: Details
Choose Desired Output
Data Input
Filtering
Final Check
Final Check
Verify inputs
Save Scenario
< Back Save Scenario & Run Cancel
Click the button Press here to verify your selections. If there are any errors, MATS will
present a message letting you know. For example, if the path to a model file is invalid —
perhaps you misspelled the file name — you would get the following error:
- Verily inputs
Press here to verity your selections...
Checking...
Model baseline data missing.
Check OK. Press the finish button to continue..
After making the necessary correction, click the button Press here to verify your
selections.
Abt Associates Inc.
333
June 2010
-------
Visibility Analysis: Details
- Verily inputs
Press here to verily your selections...
Checking...
Check OK. Press the finish button to continue..
When your window looks like the window above, click either Save Scenario & Run or
Save Scenario. Save Scenario & Run will cause MATS to immediately run the scenario.
11.4.1 Running MATS in Batch Mode
The Save Scenario button will save the scenario as a configuration file (.cfg file). The
"*.cfg" file will be saved in the .\MATS\output directory. Several .cfg files can be created
with the MATS interface and run later in batch mode. To do this, edit the default batch file
located in the .\MATS directory. The file "batchmats.bat" should be edited with a text
editor to point to the name and location of the .cfg files that will be run in batch mode.
ltraEdit-32 - [C:\Documents and Settings\btimin\Desktop\MATS-5-08\batchmats.bat]
File Edit Search Project View Format Column Macro Advanced Window Help
~ & cf Q
& H Ml I W!p I H iESElg 14
FP Go ^6l0 [D'
batchmats.bat
start /wait MATS.exe "c:\Program Files\Abt AssGciates\MATS\\output\default.cfg"
start /wait MATS.exe "c:\PrGgraiti Files\Abt Assoc iates\ HATS\ \ output\default2 . cfg"
Ul
For Helpj press F1 Ln 1, Col. lj CW
DOS
Mod: 5/23/2008 5:02: ¦49PM File Size: 163
|INS | ^
After editing the batchmats.bat file, simply run the .bat file. MATS will start and run in
the background.
Abt Associates Inc.
334
June 2010
-------
Output Navigator
12 Output Navigator
The Output Navigator allows you to load results files (i.e., *.ASR files) that you have
previously created in MATS. You can view these data in maps and in tables, or export the
data to text files that you can then work with in a program such as Excel.
To start, just click on the Output Navigator tab.
Help
Start Map View p Output Navigator
Load
Highlight file of interest and right-click t
ptions to Map, View, and E
rtthe data.
Stop Info
Click on the Load button to view your file of interest. This will bring up the Open MATS
Results file window. Choose a results file (with the .ASR extension) This will bring you
back to the Output Navigator and display the available files.
Abt Associates Inc.
335
June 2010
-------
Output Navigator
Help "
Start Map View | 0utput Navigator
|i Load "i| [
Extract All Highlight file of interest and right-click to view options to Map, View, and Extract the data.
Name
|Type
Size
- Configuration/Log Files
Configuration
Configuration
Run Log
54kb
Okb
: Log File
- Output Files
Tutorial Visibility - Forecasted Visibility - all des Monitor Network
Tutorial Visibility - Used Model Reference Cells Monitor Network
Tutorial Visibilily - Used Model Reference Cells Monitor Network
Tutorial Visibility - Forecasted Visibility Data Monitor Network
Tutorial Visibility - Class 1 Area and IMPROVE I Monitor Network
6kb
35kb
35kb
2kb
17kb
] Stop Info
The files listed fall into two categories: Configuration/Log Files and Output Files. The
Configuration File stores the assumptions used in generating your results file. The Log
File stores information regarding the version of MATS used to create the results file and
the date and time of its creation.
To examine a file, right-click on the file that you want to view. For Output Files, this will
give you three choices, Add to May. View, and Extract.
Abt Associates Inc.
June 2010
336
-------
Output Navigator
Help ~
Start Map View | Output Navigator |
Load
Extract All
|Mame
Highlight file of interest and right-click to view options to Map, View, and Extract the data.
|Type
Size
- Configuration/Log Files
Configuration Configuration
Log File Run Log
- Output Files
Tutorial Visibility - Forecasted Visibility - all des Monitor Network
Tutorial Visibility - Used Model Reference Cells Monitor Network
Tutorial Visibility - Used Model Reference Cells Monitor Network
Tutorial Visibility - Forecasted Visil
Tutorial Visibility - Class 1 Area anc Vjew
Extract
tor Network
54kb
Okb
6kb
35kb
35kb
wwm
17kb
J | Stop Info
For the Configuration File and Log File you will see two options: View and Extract.
Help
Start Map View | Output Navigator I
Load
Extract All
Highlight file of interest and right-click to view options to Map, View, and Extract the data.
| Name
I Type
| Size
t) Configuration/Log Files
RBI Config u ratpi 1
Configuration
54kb
Log File '
- Output Fileo
Run Log
Tutorial Visibility - Forecasted Visibility - all des Monitor Network
Tutorial Visibility - Used Model Reference Cells Monitor Network
Tutorial Visibility - Used Model Reference Cells Monitor Network
Tutorial Visibility-Forecasted Visibility Data Monitor Network
Tutorial Visibility - Class 1 Area and IMPROVE I Monitor Network
Okb
6kb
35kb
35kb
2kb
17kb
Stop Info
Abt Associates Inc.
337
June 2010
-------
Output Navigator
12.1 Add Output Files to Map
The Map View tab is initially empty, starting with just a blank map of the United States.
To map your results, click on the Output Navigator tab. Load the ASR file that you want
to view and then right-click on the particular Output File that you want to map. This will
give you three choices, Add to Map, View, and Extract. * Choose the Add To Map option.
Abt Associates Inc.
338
June 2010
-------
Output Navigator
Help ~
Start Map View | Output Navigator |
Load
Extract All
|Mame
Highlight file of interest and right-click to view options to Map, View, and Extract the data.
|Type
Size
- Configuration/Log Files
Configuration Configuration
Log File Run Log
- Output Files
Tutorial Visibility - Forecasted Visibility - all des Monitor Network
Tutorial Visibility - Used Model Reference Cells Monitor Network
Tutorial Visibility - Used Model Reference Cells Monitor Network
Tutorial Visibility - Forecasted Visil
Tutorial Visibility - Class 1 Area anc Vjew
Extract
tor Network
54kb
Okb
6kb
35kb
35kb
wwm
17kb
L
_J| Stop Info
This will bring you back to the Map View tab.
M HATS
Help
Start | Map View | Output Navigator
•^r l'"!) V 13 E3 Standard Layers
JiData Loaded
0
Tutorial Visibility - Forecasted'
Long:-180.79131, Lat: 61.76359 ***
Extent: Min(-159.661,2.334) Max(-fr**H51.197)
J | Stop Info
Details on how to generate a variety of maps are in the Map View chapter.
Abt Associates Inc.
339
June 2010
-------
Output Navigator
* Note that if you right-click on the Configuration File or Log File you will only see two options: View
and Extract. The Add To Map option is only relevant to the Output Files, as the Configuration and
Log files do not have a geographic component.
12.2 View Files
To view a file of interest, right-click on it, and then choose View. There are three basic
types of files available: Configuration. Log, and Output files.
Help
Start Map View I Output Navigator I
Load
Extract All
Highlight file of interest and right-click to view options to Map, View, and Extract the data.
|Type
Name
- Configuration/Log Files
j Configuration Configuration
: Log File Run Log
- Output Files
j- Tutorial Visibility - Forecasted Visibility - all des Monitor Network
j- Tutorial Visibility - Used Model Reference Cells Monitor Network
; Tutorial Visibility- Used Model Reference Cells Monitor Network
54kb
Okb
6kb
35kb
35kb
Stop Info
12.2.1 Configuration File
A Configuration File stores the choices that you have made when using MATS. A useful
feature of a Configuration File is that it is reusable. You can use an existing Configuration
File, make some minor changes to generate a new set of results, without having to
explicitly set each of the choices you made in the previous Configuration.
To view a Configuration file from the Output Navigator, right-click on the file.
Abt Associates Inc.
340
June 2010
-------
Output Navigator
~n
Help "
Start Map View | Output Navigator |
Load
Extract All
|Mame
Highlight file of interest and right-click to view options to Map, View, and Extract the data.
|Type
Size
- Configuration/Log Files
Log File
Configuration
Run Log
- Output File^
Tutorial Visibility - Forecasted Visibility- all des Monitor Network
Tutorial Visibility - Used Model Reference Cells Monitor Network
Tutorial Visibility - Used Model Reference Cells Monitor Network
Tutorial Visibility-Forecasted Visibility Data Monitor Network
Tutorial Visibility - Class 1 Area and IMPROVE I Monitor Network
Okb
6kb
35kb
35kb
2kb
17kb
This will bring up the options that you chose when generating your results.
Choose Desired Output
Point Estimates
Scenario Name
Forecast
T emporally-adjust visibility levels at Class 1 Areas
IMPROVE Algorithm
C use old version
<• use new version
* U se model grid cells at monitor
Use model grid cells at Class 1 area centroid
Actions on run completion
I? Automatically extract all selected output files
< Back
Next >
Cancel
Abt Associates Inc.
341
June 2010
-------
Output Navigator
12.2.2 Log File
A Log File provides information on a variety of technical aspects regarding how a results
file (*.ASR) was created. This includes the version of MATS, the date and time the *.ASR
file was created.
To view a Log file from the Output Navigator, right-click on the file.
A separate Run Log tab will appear.
[Name |Type | Size
- Configuration/Log Files
! Configuration Configuration 54kb
Extract |
r Tutunai viyiumiy'- Forecasted Visibility - all des Monitor Network Gkb
| Tutorial Visibility - Used Model Reference Cells Monitor Network 35kb
j Tutorial Visibility - Used Model Reference Cells Monitor Network 35kb
rTutorial Visibility- Forecasted Visibility Data Monitor Network 2kb
Tutorial Visibility - Class 1 Area and IMPROVE I Monitor Network 17kb
Abt Associates Inc.
342
June 2010
-------
Output Navigator
^ HATS
BUS
Help "
Start Map View Output Navigator | Run Log |
| Close |
»» Start MATS.exe v 2.2.3.1218 2009-09-25 12:13:46
Starting iteration 0
Starting Visibility Analysis
Loading monitor data.
Completed loading monitor data. 208.557 s.
Computing geometry
Completed computing geometry. 84.844 s.
Generating RRFs and computing forecast.
Completed generating RRFs and computing forecast. 72.502 s.
Completed visibility analysis.
Total execution time: 365.947 s.
«« Stop MATS.exe 2009-09-25 12:19:52
Click the Close button when you have finished viewing it. (The Run Log tab will
disappear.)
12.2.3 Output Files
An Output file is one of the file types within a *.ASR results file. The types of Output
Files available depend on the type of analysis (PM. Ozone, or Visibility) and the output
choices that you have specified in the Configuration File.
To view an Output file Output Navigator, right-click on a file.
Abt Associates Inc.
343
June 2010
-------
Output Navigator
This will bring up a Monitor Network Data tab. The upper left panel allows you to view
the ID and latitude and longitude of the monitors in your data — at the right of this panel
there is a scrollbar with which you can locate any particular monitor of interest. The lower
left panel allows you to view the other variables in the data.
Abt Associates Inc.
344
June 2010
-------
Output Navigator
Help
Start Map View Output Navigator | Monitor Network Data |
Tutorial 03 - Ozone Monitors — monitor data, temporally adjusted 2015
Show All | or select a particular location to see c
I type
[long
010030010
010270001
010331002
010510001
010550011
010690004
010730023
ni n-701 nni
30.498001
33.281261
34.760556
32.4985667
33.904039
31.1906565
33.553056
AOCCCC
-87.8814123
-85.8021817
-87.650556
-86.1365871
-86.0538672
-85.423117
-86.815
ojLHL
Is
Select Quantities that must be >= 0
~ b_o3_dv
~ f_o3_dv
~ referencecell
J rrf
~ ppb
~ days
Export Export this data to CSV
id
date
b_o3_dv
f_o3_dv
referencece
rrf
ppb
days
i
010030010
2004
78.0
68.8
95023
0.8825
85.0
11.0
010270001
2004
79.3
62.7
108051
0.7909
71.0
11.0
010331002
2004
-7.00
-9.00
92063
0.7642
71.0
11.0
010510001
2004
76.7
63.2
106043
0.8251
70.0
9.00
f
010550011
2004
75.0
58.0
105056
0.7738
73.0
10.0
010690004
2004
-7.00
-9.00
113032
-9.00
70.0
1.00
nm73fin?3
?nn4
77 0
60?
rnnnR?
0 7828
81 n
11 n
Stop Info
The default option is to show all of the data in the lower left panel. If, however, you want
to just view the data for a particular monitor — in this example, monitor ID = "010331002"
— use the scrollbar (if needed) and then highlight this monitor. MATS will then display
the values for this monitor in the bottom panel.
Abt Associates Inc.
345
June 2010
-------
Output Navigator
Help "
Start Map View Output Navigator | Monitor Network Data |
Tutorial 03 - Ozone Monitors — monitor data, temporally adjusted 2015
Close
Show All or select a particular location to see data.
id
type |
jlat j
| long
010030010
30.498001 j
-87.88141231
010270001
33.281261
-85.8021817
010331002
34.760556
-87.650556
010510001
32.4985667
-86.1365871
010550011
33.904039
-86.0538672
010690004
31.1906565
-85.423117|
010730023
33.553056
-86.815L
mmo-inm
I
Select Quantities that must be >= 0
~ b_o3_dv
~ f_o3_dv
~ referencecell
~ rrf
~ ppb
~ days
Export Export this data to CSV
id
date
b_o3_dv
f_o3_dv
referencece
rrf
ppb
days
010331002
2004
-7.00
-9.00
92063
0.7642
71.0
11.0
To view all of the data again, click on the Show All button.
H HATS
Help
Start Map View Output Navigator | Monitor Network Data I
Tutorial 03 - Ozone Monitors — monitor data, temporally adjusted 2015
Close
3how All j| or select a particular location to see data.
I type
010030010
30.4980011
[long
010270001
010331002
010510001
010550011
010690004
010730023
m moi nno
33.2812611
34.760556
32.4985667 j
33.904039
31.1906565
33.553056
Aocncc
-85.8021817|
-87.650556
-86.13658711
-86.0538672 j
-85.423117
-86.815
ocju.
IE
Select Quantities that must be >= 0
~ b_o3_dv
f_o3_dv
~ referencecell
rrf
~ ppb
~ days
Export Export this data to CSV
id
(date
b_o3_dv
f_o3_dv
referencece
rrf
ppb
days
n
010030010
12004
78.0
68.8
95023
0.8825
85.0
11.0
010270001
2004
79.3
62.7
108051
0.7909
71.0
11.0
010331002
12004
-7.00
-9.00
92063
0.7642
71.0
11.0
010510001
12004
76.7
63.2
106043
0.8251
70.0
9.00
010550011
2004
75.0
58.0
105056
0.7738
73.0
10.0
010690004
2004
-7.00
-9.00
113032
-9.00
70.0
1.00
F
nm73nn?3
!?nn4
77 0
fill?
mnnfi?
n 7r?r
81 n
11 0
J| Stop Info
To eliminate missing values (denoted by negative numbers in the lower panel), check one
Abt Associates Inc.
346
June 2010
-------
Output Navigator
or more boxes in the panel in the upper right of the window. For example, to eliminate any
monitors that do not have a ozone design value forecast, check the forecasted ozone design
value variable "f o3 dv". MATS will automatically drop these values. (Note that the
monitor that we previously highlighted [monitor ID = "010331002] has now dropped out
of the display.)
Help
Start Map View Output Navigator [ Monitor Network Data I
Tutorial 03 - Ozone Monitors — monitor data, temporally adjusted 2015
Show All or select a particular location to see data.
Data
id
type
lat
llong
010030010
30.498001
-87.8814123
010270001
33.281261
-85.8021817
010331002
34.760556
-87.650556
010510001
32.4985667
-86.1365871
010550011
33.904039
-86.0538672
010690004
31.1906565
-85.423117
010730023
33.553056
-86.815
ni mil nno
jioccrc 1
oir m c
Select Quantities that must be >= 0
] b_o3_dv
i—
] referencecell
] rrf
ppb
] days
Export Export this data to CSV
id
| date
b_o3_dv
f_o3_dv
referencece
rrf
ppb
days
010030010
2004
78.0
68.8
95023
0.8825
85.0
11.0
010270001
2004
79.3
62.7
108051
0.7909
71.0
11.0
010510001
2004
76.7
63.2
106043
0.8251
70.0
9.00
010550011
2004
75.0
58.0
105056
0.7738
73.0
10.0
010730023
2004
77.0
60.2
100052
0.7828
81.0
11.0
010731003
2004
79.0
63.2
99052
0.8007
81.0
11.0
F
010731005
I?nfi4
Finn
61,1
qqnRn
A 7644
80,0
mn
Stop Info
If you want to save the data, you can click the Export button and save the file. (It is
unnecessary to add an extension. MATS automatically saves the file as a CSV text file and
adds a ".csv" extension to your file name.) You can then view the file in Excel.
12.3 Extract Files
Extracting files allows you to export files from MATS and view them in another program.
This is most relevant to the Output Files (as opposed to the Configuration/Log Files),
which you may want to view and manipulate in a database program such as Excel. MATS
will generate CSV files. These are easily viewable in a number of programs.
The simplest way to extract files is to check the box "Automatically Extract All Selected
Output Files" on the initial window of the Configuration File (the same window where
you give the Scenario Name). Alternatively, you can use the Output Navigator.From
Output Navigator load the results (ASR) file of interest.
Abt Associates Inc.
347
June 2010
-------
Output Navigator
Help -
Start Map View | Output Navigator |
Load
Extract All
Highlight file of interest and right-click to view options to Map, View, and Extract the data.
I Type
Name
Size
- Configuration/Log Files
j Configuration Configuration
: Log File Run Log
- Output Files
j- Tutorial 03 - Ozone Monitors - monitor data, temporally adjusted 2015 Monitor Network
: Tutorial 03 - Ozone Monitors - counly high monitoring sites, temporally adjusted 2015 Monitor Network
81 kb
1 kb
85kb
56kb
Stop Info
One quick method to extract all of the data in the results file is to click the Extract All
button. An Extracting All screen will appear with a suggested name for the folder storing
the results. MATS will use the Scenario Name as the default folder name. If desired you
can rename the folder to whatever you desire.
Extracting ALL...
Enter Directory Name:
OK Cancel
Click OK, and then MATS will export all of the files to this folder.
Abt Associates Inc.
348
June 2010
-------
Output Navigator
r C:\Program FilesYAbt Associates\MATS\output\TutoriaI 03
File Edit View Favorites Tools Help
©E
0 ,t Search Folders J | -11 \w
Address |^) C:\Program Files\Abt Associates\MAT5\output\Tutorial 03
H G°
Folders x Name Si2e Type
w § Configuration.cfg 82 KB CFG File
f|] Log File.log 2 KB Text Document
^Tutorial 03 - Ozone Monitors — county high monitoring sites., temporally adjusted 2015.csv 53 KB Microsoft Excel Comma Se
^Tutorial 03 - Ozone Monitors — monitor data, temporally adjusted 2015.csv 80 KB Microsoft Excel Comma Se
B Tr^t Program Files
(3 Abt Associates
SHCD MATS
'i configs
1^1 data
Q help
O maps
B £) output
lr^ Example 03
ED Tutorial 03
(O Tutorial Visibility
GEHf^ sampledata
B lr!i work
An alternative is to extract individual files. Right click on the file of interest, and choose
the Extract option.
This will bring up the Extracting. Enter output file name window. By default, MATS
will generate a folder with the Scenario Name (e.g., Example 03) and export a file with the
same name used internally by MATS. You can change both the folder and the exported
file name if desired.
Abt Associates Inc.
349
June 2010
-------
Map View
13 Map View
The Map View tab allows you to further explore your results. Initially, it is empty, with
just a blank map of the United States.
Help
Start |Map View | Output Navigator
t ' ("j) ^ Standard Layers '
J Data Loaded |
, k r&*t«K mf:. -•••
g: Will
sum :
Long: -69.87549, Lat: 18.02343 ***
Extent: Min(-159.661,2.334) Max(-M4,51.197)
Stop Info
This section discusses how to:
Load data onto the map;
Choosing colors to represent the data (referred to as "plotting" in MATS);
Zoom in and out on your map;
Add and remove outlines for states, counties and Class 1 areas (these outlines are
referred to as "Standard Layers"):
Exporting maps and CSV files;
Removing data from a map.
13.1 Loading Variables
There are two ways to bring data into a map. First, you may load data into a map with the
Output Navigator. Alternatively, you can load data directly from the map view taskbar
(see next sub-section).
Abt Associates Inc.
350
June 2010
-------
Map View
To map your results, click on the Output Navigator tab. Load the ASR file that you want
to view and then right-click on the particular Output File that you want to map. Choose
the Add to Map option.
Help
Start Map View | Output Navigator I
Load
Extract All
Highlight file of interest and right-click to view options to Map, View, and Extract the data.
|Type
Name
Size
- Configuration/Log Files
Configuration
; Log File
- Output Files
j™ Tutorial Visibility - Forecasted Visibility - all design values
r Tutorial Visibility - Used Model Reference Cells - Base Data
Tutorial Visibility - Used Model Reference Cells - Future Data
Tutorial Visibility-CI1
View
Extract
4PR0VE Monitor Identifiers and Locations
Configuration
Run Log
Monitor Network
Monitor Network
Monitor Network
Monitor Network
Monitor Network
54kb
Okb
6kb
35kb
35kb
ESI
17kb
Stop Info
This will bring you back to the Map View tab.
Abt Associates Inc.
351
June 2010
-------
Map View
Start | Map View | Output Navigator
+ - +x ^ SH Standard Layers ^
JiData Loaded ||
@ ° Tutorial Visibility - Forecasted'
Long:-198.18345, Lat: 39.94200 ***
Extent: Min(-159.661,2.334) Max(-&"*H51.197)
L
Usually the next step is to plot your data. This is discussed next.
13.1.1 Loading with Taskbar
You can load data for mapping directly from the MapView tab, once you have exported
your results file (as discussed in the Extract Files sub-section of the Output Navigator
section). To start, click on the MapView tab.
Abt Associates Inc.
352
June 2010
-------
Map View
£ HATS
BEE
Long:-196.63458, Lat: 41.63867 ***
Extent: Min(-159.661,2.334) Ma<-fr"*H51.197)
Start | Map View | Output Navigator
JiData Loaded |[
si 13
Standard Layers »
Click on the Open a monitor network file button:
This will bring you to the Open Monitor Network window. Browse to the folder with the
data file that you want to load.
Abt Associates Inc.
353
June 2010
-------
Map View
Open Monitor Network
Look in: Q T utorial Visibility
- <=¦ E & n-
^9
My Recent
Documents
a
Desktop
£J
My Documents
gJ
My Computer
H
Mj,1 Network File name:
Places
Files of type:
] Tutorial Visibility - Class 1 Area and IMPROVE Monitor Identifiers and Locations,csv
I Visibility - Forecasted Visibility - all design values.csv
^Tutorial Visibility - Forecasted Visibility Data.csv
^Tutorial Visibility - Used Model Reference Cells - Base Data.csv
Visibility - Used Model Reference Cells - Future Data.csv
Tutorial Visibility - Forecasted Visibility Data.csv
[Monitor Network _»j
Open
Cancel
Click the Open button after selecting your file (or just double-click on the file you want to
load) and this will take you to back to the Map View tab.
Abt Associates Inc.
354
June 2010
-------
Map View
HATS
BMM
Help "
Start | Map View | Output Navigator
Long: -185.64031, Lat: 57.86424 ***
Extent: Min(-159.661,2.334) Max(-&*~H51.197)
13.2 Plotting a Value
To plot a value you must first load one or more variables into the MapView. Right click
on the text in the left panel and choose Plot Value.
Abt Associates Inc.
355
June 2010
-------
Map View
Help
Start [~Map View | Output Navigator
vi_
JiData Loaded i|
Standard Layers »
0 • Tutorie11'1-11-^ - lMBB1
Remove
Export as CSV File
Long: -173.21473, Lat: 28.36625 ***
Extent: Min(-159.661,2.334) Max(-&"*H51.197)
This will bring up the Shape Class Breaks window.
Shape Class Breaks
Layer Name: Tutorial Visibility - Forecasted Visibility Data
Value:
Date
2002
(* Bins C Unique Values
3
3
Class Count: [5 Marker Sizing: 0 ~U
Start Color
End Color
tig Clear B reaks
V Apply
JK Close
Here you can choose the variable (or "Value") that you want to display and how it will be
seen. Scroll through the drop-down Value menu and choose dv best. This is forecasted
visibility (measured in deciviews) on the days with the best visibility. (Note that a
description of all results variables generated by MATS are described in the "Details"
sections for Annual PM. Daily PM. Ozone, and Visibility.)
Abt Associates Inc.
356
June 2010
-------
Map View
Shape Class Breaks
Layer Name: Tutorial Visibility - Forecasted Visibility Data
I
Value:
3
Date
Class Count:
dv best
dv_worst
base_best
base_worst
rrf_b_crustal
rrf_b_no3
rrf_b_oc
rrf b ec
Start Color
End Color
There are a variety of display options that you can choose. These options are discussed in
detail in the Plotting Options section. After choosing your display options, then click the
Apply button. View the map in the Map View tab. (You can move the Shape Class
Breaks window, if it is obscuring the map.)
If you want to change your display options, go back to the Shape Class Breaks window,
make the changes, and click Apply again. You may do this as many times as needed.
When you are satisfied with the map, click the Close button in the Shape Class Breaks
window. This will bring you back to the Map View tab.
Abt Associates Inc.
357
June 2010
-------
Map View
IS MATS
B1
Help
Start | Map View | Output Navigator
V -X '
"v.* Si
Standard Layers T
JiData Loaded ;|
© • Tutorial Visibility-Forecastdu''
@ • dv_best 8.94 to 10.62
@ * dv_best 10.62 to 12.03
@ ° dv_best 12.62 to 14.01
@ ° dv_best 14.74 to 14.74
i ¦
Long:-182.66958, Lat: 58.50817 ***
Extent: Min(-159.661,2.334) Max(-&~*H51.197)
13.2.1 Plotting Options
MATS gives you a number of plotting options with the Shape Class Breaks window.
These are demonstrated with the results file "Tutorial Visibility - Forecasted Visibility - all
design values.csv" generated after completing the visibility tutorial. The same concepts
hold for other results files.
With the Date drop-down menu you can specify a particular year (assuming the data have
multiple years).
Abt Associates Inc.
358
June 2010
-------
Map View
Shape Class Breaks
Layer Name:
Value:
Date
Class Count:
T utorial Visibility - Forecasted Visibility - all design vak
|base_v
worst
"3
2002
T
2000
2001
2002
2003
2004
Start Color
End Color
iJS Clear Breaks
Apply X Close
With the Class Count option, you can specify into how many groups you want to divide
your data. The default is to use 5 bins. For most purposes this is a reasonable number.
Shape Class Breaks
Layer Name: Tutorial Visibility - Forecasted Visibility - all design vali
Value:
Date
basejworst
2002
E
~3
(* [Bind C Unique Values
Class Count: Marker Sizing: F I
Start Color
End Color
35 Clear Breaks V Apply X Close |
If you choose Unique Values option, you will have a separate bin or group for each unique
value in your data. This can lead to hundreds of bins. Generally, this is not an option that
you would want to choose. Note that if you choose this option, the Class Count and
Marker Sizing (discussed next) will be inoperative.
Abt Associates Inc.
359
June 2010
-------
Map View
Shape Class Breaks
Layer Name: Tutorial Visibility - Forecasted Visibility - all design vali
Value:
base_worst
J
Date
12002
C Bins
(* [Unique Values!
Class Coun
|5 3
Marker Sizing: [O
±j
Start Color
End Color
Clear Breaks
V Apply
X Close
The Marker Sizing allows you to vary the size of the marker on your map based on the
data values. The default is a Marker Sizing value of "0", which keeps the marker on your
map all the same size. A value of" 1" and higher gives the larger values progressively
larger markers on the map.
Shape Class Breaks
Layer Name: Tutorial Visibility - Forecasted Visibility - all design vali
Value:
| basejworst
T3
Date
12002
~3
Bin
Class Count:
u
C Unique Values
Marker Sizing:
Start Color
End Color
35 Clear Breaks V Apply X Close |
The Start Color option allows you to set the color of the markers for the lowest values.
The End Color option allows you to set the color of the markers for the highest values.
MATS uses a mix of these two colors for intermediate values. The default colors are blue
and yellow for the start and end.
If you want to change the Start Color, click on the blue square. This will bring up the
Color window. The simplest option is to click on the color you prefer from the pre-
defined Basic colors panel (in the upper half of the Color window), and then click OK.
Abt Associates Inc.
360
June 2010
-------
Map View
(You can also double-click on desired color.)
If for some reason, you do not see the color you want to use in the Basic colors panel, you
can click the Define Custom Colors button. Click in the large multi-color square to
identify the color you want and then adjust the hue with the slider bar on the far right. You
can save the color you generate by clicking the Add to Custom Colors button.
3asic colors:
nrnirirr
¦ rriiin
B
Custom colors:
Hue: 15 Red: 232
Sat: |166 Green: |166
Define Custom Colors »
ColorlSolid
Lum: [163 Blue: fl25
OK
Cancel
Add to Custom Lolors
When satisfied, click OK. This will bring you back to the Shape Class Breaks window.
Abt Associates Inc.
361
June 2010
-------
Map View
Shape Class Breaks
Layer Name: Tutorial Visibility - Forecasted Visibility - all design vali
Value:
Date
| base_worst
2002
jd
(* Bins
C Unique Values
I5 *l
Marker Sizing: [ft
±l
Start Color
End Color
~^§ Clear Breaks
V" Apply
X Close
You can change the End Color through a similar process.
To test how your colors look, click the Apply button. If you do not like what you see,
click the Clear Breaks button. When you are finally satisfied with the look of your map,
click the Close button.
13.3 Zoom Options & Pan View
In addition to the plotting options available in the Shape Class Breaks window, there are
various options on the task bar that you can choose to adjust the map. There are standard
Zoom in and Zoom out options, as well as a Pan option that lets you
manually move the map.
(*j)T
In addition, there is a Zoom to an area drop-down menu . This lets you zoom to
pre-specified regions, or "zoom frames", such as the continental US.
Abt Associates Inc.
362
June 2010
-------
Map View
Maryland
New England
Southern California
Texas
Washington DC
Edit Zoom Frames
Add Current View to List
jf~MATS
Long:-130.57694, Lat: 47.70384
Extent: Min(-122.607,21.608) Max(^.820,45.788)
Start | Map View | Output Navigator
m B Standard Layers *
If desired, you can change the "Zoom Frames" to whatever you are currently viewing.
Choose Add Current View to List from the list of options in the drop-down menu.
Add Current View to List
Start | Map View | Output Navigator
Long:-102.14858, Lat: 45.32199 ***
Extent: Min(-102.175,26.468) Max(^*.844,38.027)
"" +v K' jf|^ Standard Layers
Full Extent
Continental US
Maryland
New England
Southern California
Texas
Washington DC
Edit Zoom Frames
to 28.47
to 29.94
4 to 30.54
to 30.98
Abt Associates Inc.
363
June 2010
-------
Map View
This will bring up the Add Frame window. Type in whatever name you want to use for
this "zoom frame" and this will be available whenever you use MATS.
Add frame..
13.4 Standard Layers
The Standard Layers drop-down menu allows you to add and remove state, county, and
Class 1 area borders. By default, MATS displays the state and county borders. These can
often provide useful context to maps, however, at times they can obscure the markers
somewhat — this is most often a problem with the county boundaries.
To eliminate a layer, open the Standard Layers drop-down menu and click on the active
layer that you want to remove. This will bring up a map view with the layer removed.
SMM
Long:-99.66404, Lat: 46.22975 ***
Extent: Min(-103.909,25.321) M9x(^* 408,37.946)
Start | Map View | Output Navigator
\' jr|^ Standard LayersT |
JjData LGadeci]|
@ ° Tutorial Visibility-Forecasteu
@ ° base_worst 25.41 to 28.47
@ O base_worst 28.69 to 29.94
® O base_worst 29.94 to 30.54
@ O base_worst 30.98 to 30.98
|| | Stop Info
To add a layer back, choose the layer you want to add from the Standard Layers drop-
down menu.
Abt Associates Inc.
364
June 2010
-------
Map View
13.5 Exporting Maps & Data Files
MATS allows you to export maps and data from the Map View tab. To export a BMP file.
click on the Export current map view to an image file option 85 . (The next sub-
section discusses exporting the underlying data.)
bmm
Long:-85.81734, Lat: 40.56049
Extent: Min(-91.978,32.874) Max(-?"*193,38.181)
Start | Map View | Output Navigator
T . J jjfli Standard Layers '
J Data Loaded ||
@ ° Tutorial Visibility-Forecasteu"
© ° base_worst 25.41 to 28.47
@ O base_worst 28.69 to 29.94
@ O base_worst 29.94 to 30.54
@ O base_worst 30.98 to 30.98
J | Stop Info
This will up a window where you can name your image. Browse to whatever folder in
which you want to store your image.
Abt Associates Inc.
365
June 2010
-------
Map View
3
is ej
-
"3
Save
Cancel
Save image to...
Save in: T utorial VisibilitiJ
u
My Recent
Documents
£
Desktop
My Documents
gl
My Computer
%
My Network File name:
Places
Save as type:
BMP
Your BMP file can be easily viewed in a variety of software applications. Note however,
that this is just an image, and you will not be able to work with it in a GIS program the way
you might work with a .SHP file.
Abt Associates Inc.
366
June 2010
-------
Map View
13.5.1 Exporting CSV Data File
If desired, you can export a CSV file with the data used to generate your map. Just right
click on variable of interest in the left panel, and choose the Export as CSV File option.
j^MATS
Start | Map View | Output Navigator
]| stop Info
Long:-91.26199, Lat: 38.38608
Extent: Min(-91.978,32.874) Max(-?—193,38.181)
jDataLoaded
^^ Standard Layers T
© O
0 O base_worst 29.94 to 30.54
© O base_worst 30.98 to 30.98
Abt Associates Inc.
367
June 2010
-------
Map View
Note that this exports the same data that the Output Navigator would export. Choose
whichever approach is easier.
Abt Associates Inc.
368
June 2010
-------
Frequently Asked Questions
14 Frequently Asked Questions
This section answers questions that have arisen when running MATS.
14.1 Error: MATS will not create a folder for extracting files
If an output folder already exists MATS will return an error if you click the Extract All
button. This occurs even if the pre-existing folder is empty.
Error (X^
Unable to create directory
OK
This error can be avoided by using a different folder name or removing the pre-existing
folder.
14.2 Where is there a description of output variables?
Descriptions of the output variables are in the separate "Details" sections for Annual PM.
Daily PM. Ozone, and Visibility.
14.3 Removing Data
You can have multiple data files in a map. If you decide to remove a datafile, right click
on the variable that you want choose the Remove option.
Abt Associates Inc.
369
June 2010
-------
Frequently Asked Questions
£ HATS
BM®
Start | Map View | Output Navigator
+ x' +\ N" £i ^ Standard Layers -
|:Data Loaded j|
@ o
Tutorial Visibility-Used Moael
@ •
so4 -6 to -6
0 *
so4 -6 to -6
@ e
so4-6 to 0.6687
@ •
so4 0.8528 to 1.0176
@ ° Tutorial
0 o
© O basejA p|0t Value
@ O base_worst 29.94 to 30.54
© O base_worst 30.98 to 30.98
Long:-91.24775, Lat: 36.36290 ***
Extent: Min(-91.978,32.874) Max(-?"*193,38.181)
This will bring back the map without the undesired data.
Help
Long:-90.90679, Lat: 41.04383 ***
Extent: Min(-91.978,32.874) Max(-?°^93,38.181)
Start | Map View | Output Navigator
: ^ \ \w j|| | ^ Standard Layers -
J Data Loaded ||
0 ° Tutorial Visibility-Used Mocjtil
© • so4 -6 to -6
© • so4 -6 to -6
© ® so4-6 to 0.6687
© ° so4 0.8528 to 1.0176
Abt Associates Inc.
370
June 2010
-------
References
15 References
Chow, J.C., J.G. Watson, L.W.A. Chen, W.P. Arnott, and H. Hoosmuller (2004).
"Equivalence of Elemental Carbon by Thermal/Optical Reflectance and Transmittance
with Different Temperature Protocols", Environmental Science and Technology, 38(16):
4414-4422.
Clegg, S.L., P. Brimblecombe, and A.S. Wexler (1998). "Aerosol Inorganics Model: A
Thermodynamic Model of the System H+-NH4+-S042"-N03"-H20) at Tropospheric
Temperatures." J. Phys. Chem. 102A: 2137:2154.
Frank, Neil (2006). "Retained Nitrate, Hydrated Sulfates, and Carbonaceous Mass in
Federal Reference Method Fine Particulate Matter for Six Eastern U.S. Cities" JAWMA.
Vol 56: 500-511.
IMPROVE (2006). "Revised IMPROVE Algorithm for Estimating Light Extinction from
Particle Speciation Data", January 2006,
http://vista.cira.colostate.edu/improve/Publications/GravLit/grav literature.htm.
Turpin, B.J. and H-J Lim (2001). "Species Contributions to PM2.5 Mass Concentrations:
Revisiting Common Assumptions for Estimating Organic Mass", Aerosol Science and
Technology, 35: 602-610.
U.S. EPA (2003). "Federal Reference Method Quality Assurance Report."
U.S. EPA (2006). "Procedures for Estimating Future PM2.5 Values for the PM NAAQS
Final Rule by Application of the Speciated Modeled Attainment Test (SMAT)"
http ://www. epa. gov/scramOO 1 /guidance sip .htm.
U.S. EPA (2007). "Guidance on the Use of Models and Other Analyses for Demonstrating
Attainment of Air Quality Goals for Ozone, PM2.5 and Regional Haze", April 2007.
EPA-454/B-07-002, http ://www. epa. gov/scramOO 1 /guidance sip .htm.
Abt Associates Inc.
371
June 2010
------- |