6/25/2020	CSN & IMPROVE Protocol Assessments - How It Works
C^jCSN & IMPROVE Protocol Assessments: How It Works
Overview
In 2014, an assessment was conducted of the Chemical Speciation Network (CSN). Historically,
the CSN has consisted of the Speciation Trends Network (STN), a core set of 52 speciation trends
analysis sites, as well as a variable number of Supplemental sites. The STN sites and the majority
of the Supplemental sites utilize a national EPA contract for shipping, handling and laboratory
analysis. The recommendations from the CSN assessment were implemented in September 2014
through January 2015.
In mid-2015, an assessment was conducted of the Interagency Monitoring of Protected Visual
Environments (IMPROVE) Protocol sites. These sites are considered CSN Supplemental sites where
the monitoring agencies chose to operate them as IMPROVE Protocol sites when the CSN was
established. They are operated under the National Park Service (NPS) contract and use IMPROVE
samplers and IMPROVE's laboratory analysis contract. They were excluded from the 2014 CSN
Assessment because it only considered sites using the CSN national contract for shipping,
handling and laboratory analysis. The CSN and IMPROVE Protocol Assessments used the same
objectives-based approach and nearly identical evaluation criteria.
Goals of the assessment:
•	Create a speciation network that is sustainable going forward.
•	Redistribute resources to new or high priorities from those of low-priority or low-benefit.
•	Extract more value from the existing network.
•	Fully leverage the value of other existing networks (e.g., IMPROVE).
Network background:
Prior to the network assessments, the CSN Network consisted of 227 sites. The breakdown is as
follows:

Using CSN
National
Contract
Lab for
Shipping &
Analysis
Evaluated in
2014 CSN
Assessment
Using
Independent
Labs for
Shipping &
Analysis
Using
IMPROVE
National
Contract Lab
for Shipping
& Analysis
Evaluated in
2015
IMPROVE
Protocol
Assessment
52 STN Sites
X
X



122
Supplemental
Sites
X
X



15
Supplmental
Sites*


X


38 IMPROVE
Protocal Sites



X
X
1/5

-------
6/25/2020	CSN & IMPROVE Protocol Assessments - How It Works
*At the time of the assessments, EPA was aware of 15 Supplemental CSN sites that chose not to
use EPA's or IMPROVES national contract for shipping, handling and laboratory analysis. These
sites were not considered in either assessment process.
CSN Network (click for full size)
Alaska
Hawaii
•	IMPROVE Protocol Sites
A CSN Trends Sites (STN)
•	CSN Supplemental Sites
1 X
Puerto Rico &
Virgin Islands
\ 4	^
W '	c?
Approach
An objectives based approach was taken in an effort to optimize the network to meet the
primary objectives, which include support of PM2.5 implementation, aiding in interpretation of
health effects and exposure research studies, and the detection of trends.
We were aware that the CSN data are used for many secondary objectives (e.g., urban
increment, source attribution, etc.). However, only the primary objectives were used as the
basis for the assessment.
A two-step decision-making process was utilized to make recommendations regarding the
continued funding of shipping and analysis at the 174 CSN sites using EPA's national contract and
the 38 IMPROVE Protocol sites using IMPROVE's national contract. The additional 15 CSN sites
using independent laboratories were not considered in the assessment process.
Scoring
Step 1: Primary Objectives Scoring
2/5

-------
6/25/2020	CSN & IMPROVE Protocol Assessments - How It Works
Sites were evaluated and provided points according to the table below for meeting the primary
objectives of the network.
Parameter
Points
NCore Site
+5
Annual DV for site: > 12 Mg/m3
+5
24-hr DV for site: > 35 Mg/m3
+5
STN Site
+4
Annual DV for county > 12|jg/m3 (site <=12|jg/m3)
+3
24-hr DV for county > 35|jg/m3 (site <=35|jg/m3)
+3
Collocated with a PM2.5 Daily FRM monitor
+3
Health effects/exposure city of interest
+3
Continuous monitor(s) (i.e., OC/EC, sulfate)
+2
Collocated* with:
IMPROVE
+ 1
NATTS
+ 1
PAMS
+ 1
* Although collocation did not meet a "primary objective", it did meet an assessment goal of leveraging
with existing networks
Results:
•	The 212 sites (174 CSN/38 IMPROVE Protocol) evaluated scored points ranging from 0 to 23.
•	102 sites (73 CSN/29 IMPROVE Protocol) scored 3 points or less, and were identified for further
assessment.
•	In Step 2, a decision matrix was created for these sites to determine funding recommendations.
Decision Matrix
Step 2: Decision Matrix for 102 Sites
A matrix was created using the scoring outlined in the table below to further determine funding
recommendations for the 102 sites identified in Step 1. Positive points were applied for
parameters that added value. Negative points were applied for parameters that did not add
value.
Parameter
Range
Points (2014
CSN
Assessment)
Points (2015
IMPROVE
Protocol
Assessment)
Annual DV (if applicable)
10-10.9 |Jg/m3
+ 1
11-11.4 |jg/m3
+2
>=11.5 |Jg/iTi3
+3
24-hr DV (if applicable)
30-31 jjg/m3
+ 1

3/5

-------
6/25/2020
CSN & IMPROVE Protocol Assessments - How It Works
Parameter
Range
Points (2014
CSN
Assessment)
Points (2015
IMPROVE
Protocol
Assessment)

32-33 |jg/m3
+2
>=34 Mg/m3
+3
Population
>= 1 Million
+ 1
County Emission Rank
Top 1.1-5%
+ 1 / pollutantA
Top 1%
+2 / pollutantA
Nearest Speciation Site (CSN or IMPROVE)
>200 km
+ 1
Statistically significant increase in trend
concentration
n/a
+1 / pollutant6
Increased model bias or error if site
removedc
n/a
+1 /
pollutant0
n/a
Correlation with major species at the
nearest 5 speciation (CSN or IMPROVE) sitesE
R>0.75 for ALL
major species
-5 (regard
sit
ess of # of
es)
R>0.75 for PM2.5
-1 /site
R>0.75 for S04
-0.2/site
-0.25/site
R>0.75 for N03
-0.2/site
-0.25/site
R>0.75 for OC
-0.2/site
-0.25/site
R>0.75 for EC
-0.2/site
-0.25/site
R>0.75 for NH4
-0.2/site
n/aE
Nearest 5 speciation (CSN or IMPROVE) sites
<150 km
-1
Nearest CSN Site
<50 km
-1
A Pollutants considered using 2011 NEI Emissions: PM2.5, SO4, NO3, OC, EC, VOCs and SO2.
B Pollutants considered (>75% records above MDL and overall average precision <10% for 2010-2012):
PM2.5, SO4, NO3, OC, EC, NH4, Fe, S, K, Ca, Na, Si, Cu, Br and In.
c Only evaluated in the 2014 CSN assessment due to staff availability.
D Pollutants considered: PM2.5, SO4, NO3, total carbon and NH4.
E Major species considered: PM2.5, SO4, NO3, OC, EC and NH4 (when applicable). The IMPROVE Protocol
sites do not measure NH4, and the scoring was adjusted to account for this (5 major species worth 0.2
points per site in the 2014 CSN Assessment and 4 major species worth 0.25 points per site in the 2015
IMPROVE Protocol Assessment.)
Results:
•	The 102 sites (73 CSN/29 IMPROVE Protocol) scored points ranging from -15 to 12.
•	19 sites (11 CSN/8 IMPROVE Protocol) scoring 2-6 points were recommended for
continued funding, as well as 12 additional sites (9 CSN/3 IMPROVE Protocol), due
to various intangible reasons that were not captured in the decision matrix
scoring (e.g., potential nonattainment county).
•	After receiving feedback from the EPA regions, state, local and Tribal monitoring
agencies, the IMPROVE Monitoring steering committee, researchers, academics,
4/5

-------
6/25/2020	CSN & IMPROVE Protocol Assessments - How It Works
and at the National Ambient Air Monitoring Conference, an additional 21 sites (11
CSN/10 IMPROVE Protocol) were recommended for funding.
• Bottom Line: after incorporating scoring, intangibles, and feedback, 38 CSN sites
were defunded in January 2015 and 8 IMPROVE Protocol sites were defunded in
January 2016. Note: State agencies decided to externally fund 4 CSN sites
beginning January 2015. All sites recommended for defunding were given the
opportunity to continue monitoring should external funds (state or local) be
provided.

-------