c/EPA
            United States
            Environmental Protection
            Agency
            Robert S. Kerr Environmental
            Research Laboratory
            Ada, OK 74820
EPA/600/8-90/004
January 1990
            Research and Development
Geostatistics  for Waste
Management:

A User's  Manual for the
GEOPACK (Version 1.0)
Geostatistical  Software
System

-------
                                                EPA/600/8-90/004

                                                 January,  1990
      GEOSTATISTICS  FOR  WASTE MANAGEMENT:

 A User's Manual For the  GEOPACK  (Version 1.0)
         Geostatistical  Software  System
                       by
                    S.R.  Yates
                    USDA/ARS
            U.S. Salinity  Laboratory
              Riverside, CA 92501

                       and

                    M.V.  Yates
  Department  of  Soil and Environmental  Sciences
            University  of  California
              Riverside, CA 92521
       Interagency Agreement No.  DW12932632
                 Project Officer

                 David  M.  Walters
     Processes and  Systems  Research Division
Robert S. Kerr Environmental  Research Laboratory
              Ada,  Oklahoma  74820
ROBERT S.  KERR  ENVIRONMENTAL RESEARCH LABORATORY
        OFFICE  OF RESEARCH AND  DEVELOPMENT
      U. S. ENVIRONMENTAL PROTECTION AGENCY
              ADA,  OKLAHOMA 74820

-------
                                    DISCLAIMER

The information in this document  has  been funded wholly or in part by  the  United
States Environmental Protection Agency under interagency agreement  No.  DW-12932632
to the United States Department of Agriculture United States Salinity Laboratory.
It has been  subjected  to  the  Agency's peer and administrative review,  and  it  has
been approved  for publication  as an EPA  document.    Mention of  trade names or
commercial products does  not  constitute endorsement or  recommendation  for  use.
                                        11

-------
                                    FOREWORD

EPA is charged by Congress to protect  the  Nation's  land,  air  and  water  systems.
Under a mandate of national environmental  laws  focused  on  air  and  water  quality,
solid waste management  and control of toxic substances,  pesticides,  noise,  and
radiation,  the agency  strives  to formulate and implement actions which lead to
a compatible  balance between human activities  and  the ability  of natural systems
to support and nurture life.

The Robert S.   Kerr  Environmental Research Laboratory is  the  Agency's center of
expertise for  investigation  of  the soil  and subsurface environment.   Personnel
at the laboratory are  responsible for management  of research programs  to:    (a)
determine the  fate,  transport and  transformation rates of pollutants  in the  soil,
the unsaturated zone and the saturated zones  of the subsurface  environment;  (b)
define  the processes  to be used in characterizing  the soil and subsurface
environment as a receptor  of pollutants;   (c)  develop  techniques  for predicting
the effect of. pollutants  on ground water, soil,  and indigenous organisms;  (d)
define  and demonstrate the applicability and limitations of using  natural
processes,  indigenous to the soil  and  subsurface environment,  for  the protection
of this resource.

This user's  manual serves  the  purpose of instructing the user  in the use of
GEOPACK,  a  comprehensive,   user-friendly  geostatistical  software  system.   This
guide should  help  the  end-user,  both  novice and sophisticated,  to become  familiar
with the  features of GEOPACK.   By using  GEOPACK,  and spending  a little  time
becoming familiar with  geostatistics,  end-users  will  be able  to include  these
geostatistical  techniques  in their work and  research  environments.
                                     Clinton W. Hall
                                     Director
                                     Robert  S.  Kerr  Environmental
                                      Research  Laboratory
                                       111

-------
                                   SUMMARY

      A comprehensive, user-friendly geostatistical software system called GEOPACK
has  been  developed.   The  purpose  of  this software  is to make  available  the
programs necessary to undertake a geostatistical analysis of spatially correlated
data.  The programs  were  written  so  that they can be used by scientists,  engineers
or  regulators with  little experience  in  geostatistical  techniques and  still
satisfy  the requirements  of  more  advanced users.   Using these  programs,  and
spending a  little  time  becoming familiar with  geostatistics the end-user  should
be  able to  include  these techniques in  their work and  research environments.
Acknowledgements:

     The   research and development  of the computer  software described in  this
report was supported in part by U.S.   Environmental Protection Agency,  R.S.  Kerr
Environmental  Research Laboratory,  Ada, OK,  74820,  through  an interagency
agreement, no.  DW-12932632.    It  has not been  subject  to the Agency's peer and
administrative  review  and  therefore  may  not  necessarily  reflect the  views  of the
Agency,  and no  official endorsement  should be inferred.
Disclaimer of  Warranty

      Although the authors of this  software have  endeavored to produce  accurate,
reliable and correct software,  this software (including instructions  for its  use)
is provided  "as  is"  without  warranty, expressed  or  implied.   Furthermore,   the
authors do not warrant,  guarantee or  make  any  representations regarding  the  use,
or the results of  the use,  of the  software or written materials concerning the
software  in terms of correctness,  accuracy,  reliability, currentness  or otherwise.
The entire  risk  as to the results  and performance  of  the  software is  the  sole
responsibility of the user.  if  the software or written materials are  defective,
the user  and  not  the author,  agents  or  employees assume  the entire cost  of all
necessary servicing,  repair or  correction.


Compaq'is a registered trademark of the  Compaq Corporation. IBM*is  a  registered
trademark of the  International Business Machines Inc.  Microsoft" and MS-DOS* are
registered trademarks of  the Microsoft Corporation.   Hercules'is  a registered
trademark of the Hercules Corporation.   Zenith" is a registered trademark  of the
Zenith Corporation.  HP°,  ThinkJet" and LaserJet" are registered trademarks  of the
Hewlett Packard Company.

                                       iv

-------
                             TABLE OF CONTENTS

                                                                            Page
Foreword	    iii

Summary	     iv

Acknowledgements 	     iv

List of Figures	      vi

  Section
     1   Introduction	      1
     2   Software Description and Operation  Instructions
            1. Computer Requirements	'	      4
            2.  Installing GEOPACK
                  a.  Installation instructions  	      4
                  b.  Temporary storage directory  	  ,  	      5
                  c.  Setting or  changing the  GEOPACK directory
                      specification  	      5
                  d.  GEOPACK System Default  Settings 	      5
            3.  System Overview
                  a.  Program structure 	      6
                  b.  Function keys and cursor movement  	      6
                  c.  File-naming conventions  	      7
            4. Data File
                  a.  Format instructions 	      9
                  b.  Examples  of Data Sets	11
            5.  The  Menu  System
                  a.  Description of the MAIN Menu	13
                  b.  Description of the HELP  Menu         (Fl)	14
                  c.  Description of the UTILITY Menu      (F3)	15
                  d.  Description of the VIEW  Feature      (F4)	18
                  e.  Description of the USER  Menu         (F5)	18
                  f.  Description of the Data Set Utility Menu	20
                  g.  Description of the Statistics Menu	24
                  h.  Description of the Variogram  Menu  ,	26
                  i.  Description of the Kriging Menu	     32
                  j.  Description of the Disjunctive Kriging  Menu ....  35
                  k.  Modifying and Adding Menus  	  36
                  1.  Trouble-shooting	40
              6.  GEOPACK  Enhancements
                  a.  Running GEO-EAS geostatistics software  	  40
                  b.  Description of GEOPACK utility programs 	  40
     3   Examples
            1.  Virus  Decay Rates In Tucson  Ground  Water 	  44
            2.  Salinity In a  Southwest  Arizona  Field  	  54
            3.  Surface Moisture, Temperature and  Texture  	     60

References  and  Suggested Literature  	  66

-------
                                 LIST OF  FIGURES

Figure 3.1.1.  Example output from Variogram Calculation  (Sample) program for the
data set, VIRUS.DAT.   m A and  B, respectively, are the tabular values and a plot
of  the  semivariogram  function.

Figure 3.1.2.   Example output  from  the  Automatic Model Fit program.   The solid
line is  an  spherical  model  that was fitted to  the  sample semivariogram.

Figure 3.1.3.  Example output  from  Manual Model Fit,  Select Model(s) program  (in
A) and the  Edit Variogram Model File (in B) .

Figure 3.1.4.  Contour diagram  of the estimated value of the decay rate using the
Ordinary  [Co]Kriging  and Line  Contour Diagram programs is  illustrated.

Figure 3.1.5.  Contour diagram using the  Ordinary  [Co]Kriging  and  Block Contour
Diagram  programs.   In A and B,  respectively,  the estimated value  for the decay
rate and  the estimation  variances  are illustrated.

Figure 3.2.1.    The  semivariogram  function for  the natural  logarithm  of  the
electrical  conductivity using the Variogram  Calculation  (Sample)  program and the
data set, ECSAR.DAT.

Figure  3.2.2.     Contour diagram  of the estimated value  of  the  electrical
conductivity using  the Disjunctive  [Co]Kriging and Block Contour Diagram programs.

Figure 3.2.3.  Contour diagram  of  the conditional probability  that  the estimated
value of  the electrical  conductivity is greater than the natural logarithm of 4
dS/m  (i.e.,  1.38  log{dS/m})  using the Disjunctive  [Co]Kriging  and  Block Contour
Diagram  programs.   This figure shows  the actual screen position of the contour
levels window when  displayed.

Figure 3.3.1.  The semivariogram function for the surface  moisture content (in A)
and the  surface soil  temperature  (in B)  using  the Variogram Calculation  (Sample)
program and  the data  set, C912.DAT.

Figure 3.3.2.  The  cross-semivariogram function  for  the  surface moisture content
and the surface soil  temperature using the Variogram Calculation  (Sample) program
and the data set,  C912.DAT.

Figure  3.3.2.     Contour diagram  of the estimated value  of  the electrical
conductivity using  the Ordinary [Co]Kriging  and Block Contour  Diagram  programs.

Figure  3.3.3.     Contour diagram  of the estimation variance  for electrical
conductivity using  the Ordinary [Co]Kriging  and Block Contour  Diagram programs.
                                        VI

-------
                                  SECTION  1

                                INTRODUCTION

      To use  geostatistical  techniques in the  analysis  of spatially correlated
data generally  requires  the use of  a  computer to handle the large  number of
samples and carry out  the lengthy calculations.   Unless  someone is known who is
willing  to provide  the  necessary computer  programs,   one is faced with the
difficult  task  of  finding,  purchasing  or developing the required computer
software.    Although  there  are  a  number of practicing  geostatisticians who
undoubtedly  have access to  the necessary programs,  these  programs  are not
generally  available  or are proprietary codes.    Often,  the programs  which are
developed  for research purposes   are  subject  to limited  availability  and are
difficult  for others  to use or  modify  for purposes other  than those  for which
they were  originally designed.

      GEOPACK has been developed in  cooperation  with the  U.S.  EPA's  R.S.  Kerr
Environmental  Research  Laboratory, Ada,  OK.   It  was developed with  the  philosophy
that geostatistical  software is  needed that  can  be used  as a  learning  tool by
individuals with little  or no geostatistical  expertise and yet can also satisfy
the needs  of  individuals  with more  advanced  training  in geostatistical methods.
The specific  objectives in creating GEOPACK were  to  develop:    1)  geostatistical
software which is easy  to  use  so  that  those with little training  in  geostatistical
methods  can  learn  these  techniques and eventually use  them in their  work
environment;   2)   an integrated system which will  free the user from  excessive file
editing and program manipulation; 3)  a system which is   adaptable in the sense
that additional  programs  can  be  incorporated  into the system by the end user at
a later date  without  having  to alter previous  programs or recompile the entire
system;   4) programs  which produce graphic output in  a variety of forms and of
publishable quality  to  meet  the needs  of research  scientists and engineers and 4)
software which includes  on-line  help  facilities and extensive  error checking in
the programs.    The  on-line  help facilities  offer  information concerning the
operation of  the  system,  its  capabilities and limitations,  how to alter  the system
as well as programming  conventions and  definitions.

      GEOPACK allows the  incorporation of other geostatistical  programs,  such as
GEO-EAS  (reference),  so the  features  of this and other programs can be accessed.
Examples showing  how these geostatistical  programs  can be used in the analysis of
spatially correlated data can  be found  in Yates et  al.  (1986  a,b,c),  Yates  (1986),
Yates  and Warrick (1987),  Yates et al.  (1988)  and Yates  and Yates  (1988, 1989).
For individuals  interested in learning geostatistical techniques,  a  number of
texts are  available  and include:  Clark  (1979),  Journel and Huijbregts  (1978),
Journel (1988) and David  (1977) .

      GEOPACK geostatistical software  system is a  package  of  programs for
conducting   analyses of the spatial variability  of one or more   random functions.
The system  is menu driven, and simplified  so that a minimum number of input data
are needed.   The programs also  limit  the  amount of intermediate results  printed
to the screen or the printer.

      GEOPACK uses  dynamic  allocation of memory  so  that  data  sets  with a  wide
range  of variables and  positions  can be  used without having  to alter the program.

-------
A large storage array  (currently set  to  approximately  10000  storage  locations)  is
partitioned based on the  number  of  samples and variables so that there is  little
wasted space compared to defining the arrays to have a  fixed number of  samples  and
variable.    One  limitation is  that  GEOPACK allows data  base  (i.e. a  data  file)  to
contain a  maximum of 10  variables  plus  their x and y positions and a sample  or
position number  (the  sample  number  must  be a  real  number).   During execution  of
GEOPACK,  whenever additional storage space (i.e.  memory)  is  required  by  a  program,
the  space  is obtained  from the  large storage  array.  if attempts are made  to  use
more memory than  is available,  an error  message is printed  out  giving  the  memory
status.  From this information,  a decision can be made  on how to  reduce  the  memory
requirements to allowable limits (i.e. reducing the number of variables  or samples
considered, etc.).

       The  GEOPACK system  includes programs to do the more  common statistical  and
geostatistical  analyses.    The  system  is estimation oriented in that if the
ordering  in the menu  system  is followed,  a  grid of estimates  for  the  selected
variable in the data set  will  result. A  description of the  various  components  of
the  system follows.


Help Facilities

       The  program includes  on-line help  facilities to provide the  user with
information  concerning   the   operation of   the program,   data   requirements  ,
conventions,  definitions,   run-time  errors,   missing  files,  etc.   that are
encountered during  execution.   At  the menu level,  the  help information is of  a
general nature.    During  execution of a  program,  the help is more specific, such
as defining a term.


User-Defined  Programs

       GEOPACK includes a  feature which makes  it  simple to access any user-defined
program (i.e. program  not included  with  GEOPACK) for generating statistics, line
graphics,   surface graphics,  text editing  and  data  base  management  systems.   The
program names are defined in the GEOPACK  System Default Settings  file (see Section
2.5.c).  Other miscellaneous user-defined  programs  can also be  incorporated into
GEOPACK using the USER'S  Menu  (see  Section 2.5.e).


Basic  Statistics

       Basic statistics such as the  mean,  median,  variance, standard  deviation,
skew,  kurtosis and maximum and  minimum values can be determined for the  selected
data base  (i.e.   a file containing one or  more  random  functions  plus their  x and
y coordinates) .   Programs are  also  included for linear regression,  polynomial
regression,  Kolomogorov-Smirnov  test for distribution  and calculating  several
percentiles of  a selected data set (i.e.  for a particular random function such  as
moisture  content).  GEOPACK  can also directly  access  any commercially available
statistics package  while running  GEOPACK which  allows the user  to run more
comprehensive   statistical  analyses  contained  in a   commercially  available
statistics package.

-------
Variography

      The  sample  semivariogram,  the  cross-semivariogram or a  semivariogram for
combined random functions [i.e.  ,  Z (x)  +  Y(x)]  for a two-dimensional  spatially-
dependent random function can be  determined.  The approach used  in determining the
sample semivariogram is similar to that  outlined in Journal and Huijbregts (1978).

      A model can be fitted to the  sample  semivariogram using the nonlinear least-
squares  fitting procedure of Marquardt (1963) .   This provides  a  first  estimate
for the  coefficients to  be used  in  a  cross-validation program  and  helps  to auto-
mate the model-fitting procedure.  If the least-squares technique fails,  or other
information is  available which  should be included in the  model-fitting  process,
the traditional iterative method of manually selecting the model coefficients and
viewing a graph comparing the  sample  values  to  the model can be used.
Linear  Estimation

      GEOPACK  includes  programs to calculate the ordinary  kriging and cokriging
estimators  in  two dimensions  along with their  associated estimation  variance.
Punctual  and  block kriging and  geometric anisotropy are  included.    There is a
cross-validation option which uses the kriging estimator in a jackknifing mode to
cross-validate  the spatial correlation  structure.    It  is possible  to  include
indicator kriging  in an analysis by creating an indicator  variable  using a data
transformation program  supplied  in  GEOPACK.
Nonlinear Estimation

      Nonlinear  estimators  such  as the  disjunctive  kriging  and disjunctive
cokriging estimators can be determined along with the estimation variance and the
conditional probability that the value is  greater  than  a specified cutoff  level.
Up to 10  cutoff  levels  are  allowed.   As with the  linear estimation method,  this
type  of an analysis can be done on punctual or block support and may include
anisotropy.


Graphics  Outputs

      Various  graphics  capabilities are included  such as linear  or logarithmic
line plots,  contour and  block  (i.e.  , pixel) diagrams.   Device drivers  for  the  HP*
Laser Jet® (series II), HP plotters,  and  dot matrix  printers (similar  to Epson"
printers)  are included with GEOPACK.  Drivers for other printers and plotters can
be written and included in GEOPACK.  The graphics  programs supplied with GEOPACK
produce intermediate quality  output and are intended for quick  and easy,  albeit
rough, graphic illustrations.   For the  highest quality graphic output,  GEOPACK can
be interfaced with any  user-defined  graphics  package  so that custom diagrams can
be developed.   Using  the  USER'S Menu, any commercially available graphics packages
can be  accessed  while running GEOPACK.    A particular  data file  can  be plotted
using a commercial graphics package by either using an internal editor (if  one is
included  in the  commercial graphics  package)   or by writing  a simple format-
translation program  to  create  a  new output file in the  correct  format.

                                        3

-------
                                 SECTION 2

               SOFTWARE  DESCRIPTION  AND OPERATION  INSTRUCTIONS
2.1.  COMPUTER  REQUIREMENTS

      GEOPACK  has been written  using a combination  of Microsoft" FORTRAN and  C
programming  languages and  runs  on  IBM°-compatible microcomputers  such as  the
PC-AT,  Compaq"-286,  -386,  Zenith",  etc.  using an MS-DOS"operating  system  (ideally
version  3.30  or greater)  and 640 K  memory.    GEOPACK does  not  require a math
coprocessor but will use one  if  it  is available.   A virtual  disk  can  be used to
increase  operating efficiency if  it is  defined  to  be  the  temporary storage
directory.  GEOPACK also  requires  that  the ANSI.SYS driver be  installed for  the
screen  output  to  perform properly.   GEOPACK requires hard disk storage of about
4 Mbytes  and,  either a CGA,   EGA,  VGA  or Hercules" graphics adapter  and  the
appropriate monochrome or color  monitor.
2.2.  INSTALLING  GEOPACK

2.2.a.  Installation  Instructions,

     An installation  program  is supplied with GEOPACK to  facilitate the
installation  of  GEOPACK  onto a  hard disk.  To completely install GEOPACK,  at least
4 Mbytes  of  free space must  be available on the  hard disk.   The  installation
program can be  started  by  typing either

      A: INSTALL    destination  drive  and path>
or
      A:INSTALL

If the  optional    and  destination drive and path>  are  not  specified,
the program will provide prompts  for  this  information.   The source  drive  can  be
any legal floppy drive and the default value is A:   .   The destination drive and
path  indicates  the hard drive   and  subdirectory  for  the  "root"  of  the
geostatistical programs.  The default  destination  is  C:\GEOPACK. It  is  advisable
to specify a  destination  drive and subdirectory since a number of programs are
copied  to this  subdirectory and  could be  accidentally  deleted  if only a drive
 (i.e.,  C:) is specified,   since  typically  root  directories  are  used to store  a
variety of programs.   If  any of these programs are  deleted, GEOPACK may produce
unpredictable results.

      Before  GEOPACK  will  operate  correctly,  the  computer  must have a specific
configuration.  To install  GEOPACK,  several  modifications to  the  computer must  be
made.   These are described in  more detail  in  the  following  sections.

      1.  A temporary  storage directory must be  created.   This  is done
          automatically.
      2.  An environment variable must be set.  The AUTOEXEC.BAT file must
          be altered.
      3.  A system parameter file must  be created.   This is  done automatically.

-------
         You will  need  to  modify this file (see Section 2.5.c) to add  the
         names of  the editor,  line and contour graphics, data  base  management
         system,  etc. you  have  available.
      4e If you  have a  HERCULES  graphics  card,  you must run MSHERC.COM
         prior to  starting a GEOPACK  session.  This program should be
         put in  the  AUTOEXEC.BAT  file.
      5. The ANSI.SYS driver  (see your MS-DOS manual)  must be  installed
         in the  CONFIG.SYS  file.   This is done by  adding a statement
         like: DEVICE=C: \DOS\ANSI. SYS  to  the  CONFIG.SYS  file.   Be sure that
         the correct path  for ANSI.SYS  is  used.

2.2.b.   Temporary  Storage  Directory.

      GEOPACK uses a temporary storage  directory to hold information  such  as  the
data set,   intermediate  results,  output files,  etc.   The installation  procedure
automatically  creates   a  temporary directory  called TMP  to be used to store
intermediate results.  In general,  the temporary storage directory is transparent
to the user.
2.2.c.   Setting or  Changing  the GEOPACK Directory Specification.

      GEOPACK will operate properly only if  the environment variable %GEODIR% (the
"%"  are used to  indicate an environment variable)  is set to  the root of the
geostatistical programs.  To set this variable the following line should be  added
to the  AUTOEXEC.BAT  file

             SET  GEODIR=C:\GEOPACK

where C:\GEOPACK is  the primary subdirectory  for  the  geostatistical programs.  In
the  user-defined menus  (described in Section  2.5.d) whenever  GEOPACK sees
"%GEODIR%"   it will  substitute  "C:\GEOPACK" so that a  command:  %GEODIR%\SHOWDIR
would be  the same  as C:\GEOPACK\SHOWDIR.   Note,   the  above command can also  be
typed at  the MS-DOS  prompt  prior  to  starting GEOPACK.   After typing the SET
command given above,  if the DOS reply:  "Out  Of  Environment Space" appears, the
user should  consult  the DOS  manual  about SHELL commands.
2.2.e. GEOPACK System Default  Settings.

      The  installation  program creates a  file  that  contains the system  default
settings.    This  file may require modification  before all of the options on  the
USER and other menus become  active.   Before making any modifications,  see  Section
2.5.C.

-------
2.3.  SYSTEM  OVERVIEW

2.3.a. Program Structure,

   The initial program  structure  of GEOPACK is shown  graphically  as
                GEOPACK
DKRG
KRIG
STAT
TMP
USER
                                     VARI
                                                      DATA
                                                     - DSTAT
                                                     - GEOEAS
                                                     - UPROG
where ,   in  the above example,  the subdirectory  GEOPACK is called  the  "root" of
geostatistical programs.  The location of the "root"  directory  is  specified during
installation and can be located on any fixed-disk drive and can be anywhere along
a subdirectory "tree".     It is recommended that the root be  the first  level
subdirectory on a given drive.

      The  subdirectories:   DKRG, KRIG,  STAT,  USER and VARI, respectively, are used
to store the  programs  for:  disjunctive kriging  and cokriging  (DRKG),  ordinary
kriging and cokriging  (KRIG),  basic statistics  (STAT),  user-defined  programs
 (USER)  and variography  (VARI).   The menu entries for each  set  of  analyses  (except
for the USER  menu)  cannot be modified by  the end-user. The TMP  subdirectory is
created to  be  used  as  the temporary  storage  directory  during  installation.

     The subdirectories of the  USER directory:  DATA,  DSTAT,  GEOEAS and UPROG are
examples of how GEOPACK can  be  altered to  include  additional  programs.   The menu
entries for the USER directory and its subdirectories  can    be  modified  by the
end-user to tailor GEOPACK for  an  individual's needs.   The instructions required
to modify  the  GEOPACK USER menu are given  in Section 2.5.k.
2.3.b. The Function  Keys  and  Cursor  Movement.

      The  function keys:   Fl, F3,  F4,  and  F5  can be  used at  any menu  in the
following manner:

      Fl  -  Help  (At  menu and program  levels.  In programs other  F-keys
             also  display help information and are indicated  when  they
             are active)

      F3  -  Utility Menu

      F4  -  View  previous screen

      F5  -  User  Menu   (User-defined programs,  menus etc.
                          Makes GEOPACK adaptable)

-------
A complete description of the entries on each of  the  menus that are accessed by
the Function keys  is  given in Section 2.5,  The Menu System.

      The cursor  (i.e.,  the light bar or highlighted  selection) can be moved by
using the following  cursor control  keys:

                      Home, End,  PgUp, PgDn, Arrow  Keys

If  these  keys  are located on  the  keypad,   the NUM-LOCK key must  be turned off.
A menu  entry can  also be selected  by pressing the letter (or number)  that is
highlighted in the  desired entry.
2.3.C.   File-Naming  Convention.

      Whenever a  data set  is  selected  (this data  set can  be located in any
subdirectory on any disk)  a file is created in the temporary storage directory and
is called  "DEFAULT".   This  file  contains  two lines,  the first gives  the  selected
file name  and  the  second the origination path of  the file.  A copy of  the  selected
file is written to the temporary storage drive as well.  These two  files  (DEFAULT
and the data set)  are  required before GEOPACK will run  properly. All  the  programs
in GEOPACK read  the  DEFAULT file to obtain  the name of the data set.  The  first
part of the  data  set  name  (i.e.,  "TEST" for a data set  named  "TEST.DAT")  is  used
for many of  the intermediate file names.   For  a data set named  "TEST.DAT" which
is assumed to contain three variables moist  (variable #1) ,  temp  (variable #2)
and sand   (variable  #3),  the typical file  names that reside  in  the temporary
storage directory  and  a brief description of their purpose follows.
       STACKER.

       MENUTEMP.BAT

       DEFAULT.



       TEST.DAT
       RPLOT.TMP
       GRAPH.DBF
       TEST.V
- Used by the menu program.

- Used by the menu program.

  Contains  the name of the data  set  and  path of
  origination.

- Data set.

  Basic descriptive statistics for all  data
  variables  in the data set.

- Temporary file  containing the  data  for the
  last regression or  least-squares graph.

  A command file  for  the graphics program.   Contains
  the commands used to generate the last  graph  (regression,
  least-squares,  contour plot, etc.).

- The  default  settings  file  for  the  semivariogram
  program.   This  file  is generated  when  the Default
  Settings  program  (on the variogram  menu)  is executed.

-------
VPLOT.DEF
VAR11.00
VAR11.01
VAR12.00
MVAR11.00
MVAR11.01
MVAR12.00
TEST.VAR
TEST.K
KPLOT.TMP
KPLOT1.TMP
TEST.D
TEST.DCi
DPLOT.TMP
DPLOT1.TMP
Temporary  file  containing the data  for  the
last  semivariogram or  cross-semivariogram graphed.

Contains  the sample semivariogram  values for
variable  1  (i.e.,  MOIST).  First  saved  file.

Contains  the sample semivariogram  values for
variable  1  (i.e.,  MOIST).  Second  saved file.

Contains  the sample cross-semivariogram values
for variable  1  with variable 2  (i.e., MOIST  vs.
TEMP).    First saved file.

Contains  the model coefficients for the sample
semivariogram values  contained in  VAR11.00.

Contains  the model coefficients for the sample
semivariogram values  contained in  VAR11.01.

Contains  the model coefficients for the sample
cross-semivariogram  contained in VAR12.00.

 Contains the semivariogram  and  cross-variogram
coefficients  for  use  by the kriging and dis-
junctive  kriging  programs.

The default settings file for the  kriging program.
This file  is generated when the Default  Settings  program
 (on the  kriging menu)  is executed.

Contains  the position, the estimate and the
estimation  variance  from the kriging  program.

Temporary  file  containing the graphic data for
the last  kriging  contour graph produced.

The default settings  file for the  disjunctive kriging
program.   This file is  generated  when the Default  Settings
program  (on the disjunctive kriging  menu)  is  executed.

Contains  the  input data required by the  dis-
junctive  kriging  program for variable  "i",  where
"i" is the  variable  number.   This  file  is
created  by  executing  the Hermite  Coefficients
program on  the  disjunctive  kriging  menu.

Contains  the position, the  estimate,  estimation
variance  and  conditional  probabilities  (if any)
from the  disjunctive  kriging program.

Temporary file  containing a contour
representation  of  one  of  the output data

-------
                         types contained in  DPLOT.TMP  (i.e.,  it has the
                         last disjunctive kriging  graph  produced).
2.4. DATA  FILE

2.4.a. Format  Instructions.

     To operate GEOPACK,  a  data file must be created in  the  proper  format.  Using
a text editor or  word processor,  a  data  file  can be entered or modified  for  use
with GEOPACK.   For GEOPACK to function  properly,  the data  set  must contain at
the minimum: 1) one   (or more)  spatially-dependent random variable(s)  (i.e.,  soil
temperature, moisture,  hydraulic conductivity,  concentration, etc.)  and   2) an
x and y position  for  each value of  the  random variable (s).   A data  file  must be
created  in ASCII  format   using   the following  standard FORTRAN formatting
instructions:
   RECORDS 1-3:
   RECORD
            FORMAT(A7 6/A76/A76)

TITLE (3)     Three  lines  of  title.


            FORMAT(2I5,F10.3)

IVAR        The  total  number of random variables  in
            the  data  set.  These are values of
            the  parameters  that will be used  in
            obtaining estimates,  etc.

NDAT        The  total  number of positions  (i.e.,
            X  and  Y  coordinate   pairs)  in the file.

CUTOFF      The  value given to this variable  is  the
            maximum allowed  value  for  the data.    Using
            the  CUTOFF,  missing data can be  excluded
            from any analysis.   To accomplish this,
            set  any  missing data in the data  file  to  a
            large  number (e.g.  99999.99).  This  number
            must be  larger  than the numeric value  of
            any  variable.   To  exclude  these large  num-
            bers  ,   set  the  CUTOFF value to a  number
            slightly  SMALLER than the missing data
            number  (e.g.,  99999.00).    Then the  mis-
            sing data will  not  be used by any program.

-------
RECORD
                 IINX
                 IX
                 IY
            FORMAT(1315)

            The  column in the data  file  that contains
            the  sample number.    if  no  sample is avail-
            able,  the program can be made  to read a
            blank  field as zeros.   In  general,  this
            will  not cause any problem in  running the
            programs.

            The  column in the data  file  that contains
            the  X coordinate data.    See  examples
            given  below for more information.

            The  column in the data  file  that contains
            the  Y coordinate data.    See  examples
            given  below for more information.
                 IV(1)       The column  in  the data file that  contains
                            one of the  random variables.   The sel-
                            ection of  the  first,  second, etc. random
                            variable is arbitrary.   The easiest  way
                            to specify  these  is to sequentially  num-
                            ber the  random variables  in the order
                             (from  left  to  right)  that they appear  in
                            the data file.    The program uses  the order
                            you specify here,  as  the  order the vari-
                            ables will  be  listed  out  on the screen.
                            If you want to change the order,  then
                            give IV(1)  as  the 1st variable, IV(2)  as
                            the second  variable,  and  so  on.
                 IV(IVAR)    The column  in  the  data file that  contains
                            IVAR-th random variable.
RECORD
            FORMAT(10(4x,A6))

NAM (I)       The  variable name for the  I-th  variable.
            NOTE:  the  I-th variable is  the  variable
            associated  with IV(I)  listed above.   See
            examples  given below for more  information.
                                     10

-------
   RECORD    7 :
                     FMT
    RECORD 8-end:

              DATA VALUES
           FORMAT(A3 0)

            Format  specification  for  the data file.
            This  should  be  of appropriate form to
            read  the data in  the  file.
                                e.g.
                                         (F5.0,2F10.3,5X,1E10.2)
              (see RECORD 7)

          The data  set  should have a sample number
           (or  index),  X coordinates, Y coordinates
          and the property(s)  to  be kriged.   Data  is
          read until  an end-of-file marker is found.
          Therefore,  there  should not be  any blank
          lines in  this part  of  the data file.

          Note: The sample no.  is assumed to be real.
2.4.b.   Examples Of Data  Sets.
   This is a typical data  file.   There are 4  random variables:  MOIST,
   TEMP,  SAND and OlL-%.   There are 119 positions  where  data was col-
   lected  (only 4 positions  are shown).
     4  119
     1235467
      MOIST     TEMP       SAND     OIL- %
  (G5.0,12F10,3)
     1     6.0000     7.0000   46.8500  999.9990    56.5102    6.5362
     2     6.0000    10.0000   46.2900    5.9250    55.6444    5.2454
   118   24.0000
   119   22.0000
21.0000   46.3500  999.9990    54.4012     4.0463
24.0000   47.1400  999.9990    52.5845    2.5345
                                        11

-------
Notes:

     1) In the above example,  the TEMP,  MOIST,  SAND and OIL are  in  columns  4,5,6
and 7,  respectively.   However,  if you note line 5 and 6 you'll see  that  MOIST  is
specified as variable  #1,  followed by TEMP,  SAND and OIL. This changes the  order
for the IV(i) and NAM(i)  data.

     2) Since  all  the random variables were  not  sampled at every location,  some
method for delineating missing data  is  needed.   In the  above example,  missing
data is specified by  entering a large number  (i.e.,  999.9990). When  executing  a
program in GEOPACK,  it  will ask you the  CUTOFF value  (for  delineating  missing
data)  and you should give a value slightly smaller than 999.9990 but  larger than
any valid data value,  for example,  999.  or 900.
   This example data  file  is  the same as the  one  above  except there are
   no sample numbers.   To  alleviate this problem  the  first  column of the
   data file is specified  as  the sample number  (i.e., they  are  all  0).
    4  119
                         6         7
                          SAND     oil-%
   1235^
   MOIST      TEMP
(G1.0,F9.3,12F10.3)
   6.0000     7.0000
   6.0000    10.0000
                         46.8500  999.9990    56.5102     6.5362
                         46.2900    5.9250    55.6444     5.2454
    22.0000
             24.0000
47.1400  999.9990
52.5845
2.5345
                                        12

-------
2.5. THE MENU  SYSTEM

     During installation,  you are asked for the name you want to use to start the
geostatistical  programs,  where the default name is GEOPACK.  After  starting the
program using the selected name,  the first menu displayed is the MAIN menu.
2.5.a.   Description of the MAIN Menu.
                          Thu  Jun  29,  1989 07:47:45 am
  Return to MS-DOS, with prompt  to  save  settings  and files
     GEOSTATISTICAL PROGRAMS: Main  Menu
  Quit GEOPACK
  Data Set Utilities Menu
  Statistics Menu
  Variogram Menu
  Ordinary  [Co]Kriging  Menu
  Disjunctive   [Co]Kriging  Menu
Esc: quit
Fl: help
F3: util
F4 : view
F5: user menu
t
Home
End
From the main menu you  can  select one of several other menus using the cursor  keys
to move  the  light bar  (or highlighted region)  over the desired selection or  by
typing the highlighted  character of the desired selection.  The  function  keys  will
display menus  intended to  offer help,  utilities  and user-defined applications.
Selecting an item from  the  MAIN  menu will produce another menu containing programs
or  other  options which are available  for undertaking  the statistical and
geostatistical analyses.    A  description of  the available menus follows.

      When leaving GEOPACK (i.e.,   typing Q  at the main menu)  another menu  is
produced which allows  you to return  to GEOPACK,  return  to  MS-DOS leaving the
intermediate data in the temporary storage subdirectory, return  to  MS-DOS  saving
the intermediate  results  to  a packed file (see Section 2.5.f)  or return to MS-
DOS and  delete the  intermediate results  from the  temporary  subdirectory  (see
Section 2.5.f) .
                                        13

-------
2.5.b.   Description of the HELP Menu  (Fl) .
  Return to Main Menu
                          Thu  Jun 29,  1989 07:47:45 am
     GEOSTATISTICAL PROGRAMS:  Main  Menu
  Quit GEOPACK
  Data Set Utilities  Menu
  Statistics Menu
  Variogram Menu
  Ordinary  [Co]Kriging  Menu
  Disjunctive  [Co]Kriging  Menu
                                                GEOPACK USER'S  MANUAL
Return To Main  Menu
Overview Of  System
Installation  Instructions
File  Formatting  Instructions
Basic System Operation
Menu Help
Utilities in GEOPACK
Adding User  Programs  and  Menus
Esc: quit
Fl: help
F3: Util
F4 : view
F5: user menu
*
Home
End
      If the Fl key  is  pressed  from any menu,  the HELP Menu is activated. On the
HELP  Menu there  are several  help files  you can access.     In general,  the
information in the User's Manual and some theory of geostatistics can be  found on
the HELP menu.
                                        14

-------
2.5. c.   Description of the UTILITY Menu  (F3).
                          Thu  Jun 29,  1989 07:47:45 am
  Return to Main Menu
     GEOSTATISTICAL PROGRAMS: Main  Menu
  Quit GEOPACK
  Data Set Utilities Menu
  Statistics Menu
  Variogram  Menu
  Ordinary  [Co]Kriging  Menu
  Disjunctive [Co]Kriging  Menu
          Utilities  Menu
Return To Main Menu
Change System Default  Settings
DOS Shell
Execute an MS-DOS  Command
Graph Most Recent  Graphic
Program Structure
Esc: quit
Fl: help
F3: Util
F4 : view
F5: user menu
t
Home
End
      If the  F3  key is pressed from any menu,  the  UTILITY  Menu  is activated.  On
the UTILITY Menu there  are several programs available  to alter basic  GEOPACK
parameters,  run MS-DOS commands while inside GEOPACK and view the current GEOPACK
program structure.   Each utility  is described in more detail below.


Change System  Default  Settings:

      This Utility  allows you to change the  default settings which  are  used  by
GEOPACK.    The default settings  include the path  specifications, device names,
screen type,  default  colors  and  names  of default  (i.e.,  user  defined)  programs.
Two files are used:  SYSETUP and SYSDEFLT  for changing  and storing  the system
default  information.   SYSETUP is the basic  file which  MUST  be present in the
GEOPACK root  directory.   This file  contains the information which  is written  to
the screen by  this  Utility  program,  an  example  is:
                                        15

-------
                           System Default Settings File
  RAM-Disk or  Temporary Directory
  Data  Storage  Directory
  Text  Color Number  (0-15)
  Border Color  Number  (0-15)
  Screen Mode  (0=Black  & White,  l=Color)
  Statistics  Program
  Word  Processor or  Text Editor
  Line-Type Graphics  Program
  Contouring Graphics  Program
  Data  Base Management  Program
                                                C:\GEOPACK\TMP
                                          C:\GEOPACK\USER\DATA
                                                             11
                                                              6
                                                              1
                                                     MICROSTAT
                                                            EOT
                                                       GRAPHER
                                                         SURFER
                                                      MFOXPLUS
  Esc: quit
Fl:  help
F2:  save
Home
End
Page Up
Page Down
Since paths  are not given for any  of  the Default programs,  it  is  required that
these programs reside in sub-directories on the current MS-DOS PATH  (so that they
can be found  from any sub-directory by  using  the  program  name only).   It is also
possible to  specify the entire MS-DOS path and program name.   PLEASE NOTE: the
programs MICROSTAT,  EDT, GRAPHER, SURFER  and  MFOXPLUS are commercially available
programs and are not included with the GEOPACK system. The Utility program writes
out a file (called SYSDEFLT which is used by the GEOPACK programs.   An example of
this file is:







Columns :
C:\GEOPACK\TMP
C : \GEOPACK\USER\DATA
11
C
b
i
MICROSTAT
EDT
GRAPHER
SURFER
MFOXPLUS
1 1
1 10







1
20
This file is used  by GEOPACK during operation of  the  programs.   At  any time if
there  is  a  change  in the definition of any item  in  SYSDEFLT a change  needs to
be made to  this  file.   The easiest way to do  so is to  use the Change  System
Default Settings option on  the  Utilities  Menu  (i.e.,  F3  key).
                                        16

-------
     To manually run the Default Settings  program type: DEFLT  sysetup  sysdeflt
from the  root  directory  of  GEOPACK.   This may be  necessary  if  the sysdeflt file
is  ever accidentally  deleted  or  modified.
DOS Shell:

      This Utility  allows  you to temporarily exit the GEOPACK  system and return
to the MS-DOS command-line mode.  When you exit to a DOS shell,  the text color on
the screen will be red.   This is to remind you  that  you  are  in  a DOS shell.  Once
you are finished running MS-DOS  commands you can return to GEOPACK by typing EXIT,
which exits you from the DOS  Shell.

     Please Note:  While  in a  DOS  shell,  some  of the  available computer memory is
still allocated to GEOPACK and if you try to run a program that requires a lot of
memory you may see the message "not enough memory".   If you must run this program,
you should exit the DOS shell  (i.e.,   return to  GEOPACK)  then exit GEOPACK.   This
will free up all  the available  computer memory.   Also,  it  is possible to exit to
a DOS  shell  and restart GEOPACK  (this  can  be done  several  times).    If this is
done,   the computer may stop operating since  there is insufficient memory to
contain both the  current  and  pre-installed  versions  of  GEOPACK.
Execute an MS-DOS Command:

      This Utility  is  similar  to  DOS Shell except that it allows you  to  run ONE
MS-DOS command.   Once you have  run the MS-DOS  command you  are returned  to the
Utility Menu.   If  you  need to run  several MS-DOS  commands,  it  is  more efficient
to exit to a DOS shell.

     Please Note:  There is slightly LESS available computer memory when Executing
an MS-DOS  command  compared to  exiting  to  a DOS shell.
Graph Most  Recent Graphic:

      This  convenience feature allows you to  replot  the  last  graph displayed (or
that  should have been displayed)  on the  screen.    This  feature is useful for
computers  with insufficient memory to display graphs  while in an application
program (i.e.  the  computer  doesn't  have  enough extra memory  to  run  the graphics
program).    Use this  options  if you don't  see a  graph and  receive the  error
message:  OUT OF MEMORY.
Program Structure:

      This  Utility gives  a pictorial representation  of the  current  GEOPACK
structure  (for  an example  see  Section  2.3.a).  Changes in the structure of GEOPACK
will be reflected in  the diagram  produced by  this program.    The USER menu,
described in Section  2.5.d, can be modified  to include any program, or add another
menu.    In  the  above  example,  the USER menu is  used to access four other menus,

                                       17

-------
DATA, DSTAT, GEOEAS  and  PROG.   These menus are  for,  respectively,  available data
sets, default  statistics program,  GEO-EAS geostatistical software and user-defined
programs.    Additional  menus and programs  can  be added by  following  the  instruc-
tions on  modifying the USER MENU given  in  Section  2.5.J.


2.5.d.   Description  of the View Feature.   (F4)

      If  for some reason you are located at one of the menus  and you need to see
the  last  screen of  output  from a program  that  has just  been completed,  you can
toggle between the menu  screen and the program  screen by  typing  F4  provided your
monitor and graphics card allows multiple  screens.
2.5.e.    Description  of  the USER Menu.   (F5)
  Return  to  Main Menu
                          Thu Jun  29,  1989  07:47:45 am
     GEOSTATISTICAL  PROGRAMS:  Main Menu
  Quit  GEOPACK
  Data  Set  Utilities Menu
  Statistics  Menu
  Variogram  Menu
  Ordinary  [Co]Kriging  Menu
  Disjunctive  [Co]Kriging Menu
             User  Menu
Return To Main  Menu
EPA GEO-EAS
Graphics Menu
Statistics  Package     (Default)
Data  Base  Management  (Default)
User  Programs
Esc: quit
Fl: help
F3: Util
F4 : view
F5: user menu
t
Home
End
       If  the  F5  key is pressed from any menu, the USER'S Menu  is  activated.   The
USER'S  Menu allows GEOPACK to be modified  to include other  programs and  menus
which are defined by the user.  The  instructions which  tell  GEOPACK which program
to run or menu to display are in a file called: USER.MEN, which is located in the
root directory of the  geostatistical  programs  (i.e.  ,  if the  default directory was
chosen,   it  would be  C:\GEOPACK).   The necessary  information for modifying the
GEOPACK  program is given under  the heading  "Adding  User  Defined Programs and
Menus" on the User's Manual Help  screen  (i.e.  , type Fl) .
                                        18

-------
      A brief description of the menu options which are  provided  with  the  original
GEOPACK system  follows.
Statistics Package  (Default):

      This feature allows you to run a default statistics package during  a GEOPACK
session.    The default  statistics package  is defined  (i.e.,  the path  and name
specified by  using the  system default setting program).   To  start this  program
type F3 and select  "Change  System Default  Settings".   In the present  example,  it
is assumed that the GEOPACK data set will have to be  modified to  a format  that  the
default  statistics package  can read.    Therefore,  selecting  this  option will
produce another menu which will let you create a file in the proper format  (Note:
this program  may have to be changed for other  statistics  programs)  and  run  the
statistic program.   The  default statistics menu  is  located in  the subdirectory:
%GEODIR%\USER\DSTAT under the  file name: DSTAT.MEN.
Graphics Menu:

      This  feature allows you to  run  one of several graphics  programs during  a
GEOPACK session.  When this option is  selected another menu is  produced which has
three options.  The first is to execute the GEOPACK graphics program and the  other
two  are  the  default  line and contour  graphics programs.


Data Base Management  System (Default):

      This  feature allows you  to run a default data base  management system during
a GEOPACK session.
GEO-EAS:

      This  feature  allows  you to run GEO-EAS during  a  GEOPACK session.

Available  Data Sets:

      This  option allows you  to  select a  data set  from  a  subdirectory  used  solely
for storing data sets.  The  subdirectory  specification  is  %GEODIR%\USER\DATA  and
the program SDATA.EXE  is used for displaying and selecting the data sets located
in  the  directory  (for  additional information  see Select  a  Data  Set  in Section
2.5.f) .
                                        19

-------
2.5.f.   Description of the  Data  Set  Utilities Menu.
  Return To Main  Menu
                          Thu  Jun 29,  1989 07:47:45 am
     GEOSTATISTICAL PROGRAMS:  Main  Menu
           Data  Set  Utilities  Menu
   Q - Return To  Main  Menu
   Select a Data  Set
   Modify an Existing  Data Set
   Pack Temporary  Directory  into  File
   Unpack Temporary Directory  from File
   Extract a File  From a  Packed File
   List Temporary  Directory  and View File
   Edit a Data Set
   Delete All Files in Temporary  Directory
Esc: quit
Fl: help
F3: util
F4 : view
F5: user menu
t
Home
End
      If  the  highlighted  key "U" is typed or  the  light  bar placed over the menu
option "Data Set Utilities Menu"  on the MAIN Menu and the enter key pressed, the
Data Set Utilities  Menu is'  activated.   On this  menu  there  are several utilities
which can be  used  to select, edit  or modify an  existing  data  set,  to pack the
contents  of  the temporary  directory into an  ARCHIVE file for later  use  or to
extract all or part  of  the  files from  the archive.
Select a Data Set:

      This program allows the selection of  an  existing data set.   The data set
must be properly formatted for use with GEOPACK.  To create a  new file,  see the
editing program  described  below.
                                        20

-------
                           Select a Data File
  C:\GEOPACK\USER\DATA\..
   .A
  C912.DAT
  ECSAR.DAT
  TEST.DAT
  VIRUS.DAT
  ESC: exit
Fl:  help
F2:  list
Pg Up
Pg Down
Home
End
      A data  set  can  be selected by either  1)  typing in the file name  or  2)  by
moving the light  bar  over  the  desired  data set and typing the ENTER key.   Other
subdirectories  can  be accessed in a similar manner  to  data  files.   Please note
that the  symbol  ..\  above indicates the  subdirectory  C:\GEOPACK\USER  and if  a
directory entry is  preceeded by a  "\",    then  the  entry  is  a directory
specification.

      This program copies  the requested data  from its current path to the
temporary directory,  creates  a  file  called  DEFAULT  which contains the program name
and its origination path.  GEOPACK uses  the  DEFAULT  file  to  determine  the proper
name of the data file.   Both  files  (i.e,  the  data file and DEFAULT) must exist  on
the temporary directory  before  GEOPACK  will  work properly.

      GEOPACK requires  that  a  cutoff value  be supplied during  execution of the
programs.   The  cutoff value  is  used to indicate missing data.   The  cutoff  value
is a specified value that is  larger  than any  of the "true" data  in the  data  set.

      For example,  if  you  have  a variable that is missing a  value at  a  point  in
space,  in  the data  set  you place a  value which is larger than any of  the  "true"
data and larger than  the cutoff  value  which  you must specify.   During execution
of a program you will give  a  value (called the cutoff value)  which is  used by the
program to determine which values to  ignore.   Any value  in the  data  set that  is
greater than  the  cutoff value  will be  ignored by the program.   It is  imperative
that the cutoff value you  specify is appropriate  for  given data  set;  if  not, the
resulting calculations will  be  inaccurate.
                                        21

-------
      For  example:  given the following data:

    index        x        y     variable 1         variable  2      variable 3
     3        10.0     11.2     12.55              45.66            1002.321
    15        14.9      1.9     11.83              48.19            1032.209
     5        18.7      7.7     12.91              36.98            1200.500
     7         5.2     14.5      9.50            9999.99            1002.321
    45        46.4     16.4     16.82              45.66            1002.321

the value used  to  indicate missing data  is  9999.99.   Note  that  this value  is
greater  than any value  of ALL the variables.    An appropriate  cutoff value  to
input into GEOPACK  would  be greater than 2000,  since  1200.5  is  the  largest  "true"
data value.
Modify an Existing Data  Set:

      This program allows  the  creation of new variables in an existing data set.
The data set must reside on the temporary directory prior to starting this program
and be properly  formatted  for  use  with GEOPACK.   The newly created  variables  can
be added to the  file.   For example,  assume  the  data set has one variable  called
the moisture content.   If a new variable log(moisture  content)  is  to be  created
it can be added  to the  file.
Pack Temporary  Directory  into  File:

      This  program allows the archiving of the files  residing in the temporary
directory.   This  is used if  you want  to  store  the  intermediate  results  from
GEOPACK for use  at a later time.   Each file is "PACKED"  into a new file in the
selected directory  under a name you provide.   If a file with the same  name  exists,
you will  be asked if you want to overwrite it.    The screen  is  similar to the
screen  shown in  the  section  "Select A Data  Set" described  above.
Unpack Temporary Directory  from  File:

      This program  allows  you to retrieve  the  files  that were previously  packed
using the program  described above.   If you  want to retrieve  only a few of the
files then use  the  program described below under  EXTRACT  A FILE.   The  screen  is
similar to the  screen  shown in the section "Select A  Data Set" described above.
Extract a File From a  Packed  File:

      This  program allows you   to retrieve one  or more of  the files that were
previously packed  using  the  PACK program described  above.   Please note that  you
must retrieve the  DEFAULT file if you  plan  to  use GEOPACK.    If  you fail to do  so
the program will print an error  message  alerting  you to  this error (the program
does not explicitly state that  the  DEFAULT  file  is-missing, however)  . The  screen
is similar to the  screen shown in the section  "Select  A Data Set" described  above.
                                        22

-------
List Temporary Directory and View  File:

      This program allows you  to  list  all  the files that currently reside in the
temporary directory.   To "VIEW" a file  place the light  bar  on  the desired file and
type return.   Also,  by typing the F2 key, additional  information  about the file
is available.  The screen  is  similar to the  screen  shown  in  the  section "Select
A Data Set"  described  above.
Edit a Data Set:

      This program allows you to edit a file.   Any file can be edited whether it
exists or not.   Once this program  is  started,  you will see the directory listing
of the %GEODIR%  directory (i.e.  ,  the root of the  geostatistical programs) and you
can select  a  file from this directory,  supply an alternate  file  (and  path)  or
give a new  directory name  to  see the listing of  the new  directory.
Remove All Files  in  Temporary Directory:

      This  command  allows you  to remove   all    the files  in the temporary
directory.   PLEASE  NOTE:  THE TEMPORARY  DIRECTORY  SHOULD  BE USED ONLY  FOR THE
GEOSTATISTICAL PROGRAMS SINCE  THIS  COMMAND WILL REMOVE  ALL  PROGRAMS  IN THIS
DIRECTORY  (or  subdirectory).

      Before you  can select a  second data  set to work on, the temporary directory
must be cleared of a previous  data set.   If  you want to save the intermediate or
final results  from this  previous  data  set,  use the  PACK  data set program described
above.
                                        23

-------
2.5.g.   Description of the  Statistics  Menu.

  ^^^^^^=^^^^^^= Thu  Jun 29,  1989 07:47:45 am
  Return To Main  Menu
     GEOSTATISTICAL PROGRAMS:  Main  Menu
                 Statistics  Menu
   Return To Main  Menu
   Sample Statistics
   Regression Analysis
   Kolomogorov Test  of  Distribution
   Least Squares  Polynomial
   Percentiles of  Data
   Variogram Menu
   Ordinary  [Co]Kriging  Menu
   Disjunctive  [Co]Kriging  Menu
Esc: quit
Fl: help
F3: Util
F4 : view
F5: user menu
t
Home
End
      If the  highlighted  key "S" is typed or  the  light  bar placed over the menu
option  "Statistics Menu"   on the MAIN Menu  and the  enter  key pressed,  the
Statistics Menu is activated.  On the  Statistics  Menu  there are several programs
from which you  can choose.   These programs  are intended  to  do the  most  basic
statistical  analyses  and are not intended  to  replace comprehensive  statistical
packages which are available commercially.  Since  it  is  anticipated that  there will
be a need for accessing a more complete statistical package, GEOPACK has built in
capabilities  to call   (i.e.,  run) a  user-defined statistical program  from the
USER'S MENU  (access to the  USER'S MENU  is obtained by  typing F5).  Since the data
format  requirements  of GEOPACK may be  different from the  commercial  statistics
package, an interface maybe necessary to rewrite the  data file in the appropriate
format.  When  this is  necessary,   it  is advisable  to have  the Default Statistic
Package option on  the  USER  menu  call  up another  menu which  has the interface and
start commands on  it.   An example is provided in the  file  %GEODIR%\USER.MEN.


Sample  Statistics:

      This program calculates the descriptive  statistics  for the  data set.  These
statistics include:   the  number  of  samples used, the mean,  median, variance,
standard deviation, skew, kurtosis,  maximum and  minimum  values of  the data  set.
                                        24

-------
Regression Analysis:

      This program calculates the linear regression coefficients, A and B, for the
linear model

     variable (i)  =  A + B*variable(j),     j  * i

A plot of the  data  and regression line is provided as  well  as the coefficient of
determination,   and  a comparison between the estimates  and data values.


Kolomogorov Test  of  Distribution:

      This program calculates the  Kolomogorov-Smirnov test statistic  for a normal
distribution.   Other distributions can  be  tested for  by transforming  the variable
of interest using the  Modify a  Data  Set command on the  DATA  SET UTILITIES MENU.


Least Squares  Polynomial:

      This program  calculates the nonlinear  regression  coefficients  C(i)  for the
polynomial model

      variable(i)  =  C(l)  + C(2)xvariable(j) +  C(3)xvariable(j)z  +  ...

A plot of the  data and regression line is provided as well as a  comparison between
the  estimates  and data values,
Percentiles of Data:

      This  program  sorts the data  set and gives  several  pre-defined percentiles
of the data.
                                        25

-------
2.5.h.   Description of  the  Variogram Menu.

                    = Thu Jun 29, 1989  07:47:45  am
  Return To  Main  Menu
     GEOSTATISTICAL  PROGRAMS:  Main Menu
                 Variogram  Menu
   Return To  Main  Menu
   Set  Program  Parameters
   Variogram  Calculation  (Sample)
   Automatic  Model  Fit
   Manual Model  Fit,  Select  Model(s)
   Edit Variogram  Model  File
   Statistics  Menu
   Ordinary  [Co]Kriging  Menu
   Disjunctive  [Co]Kriging  Menu
Esc: quit
Fl: help
F3: Util
F4 : view
F5 : user menu
t
Home
End
      If  the  highlighted key "V" is  typed  or  the light bar placed over  the  menu
option "Variogram Menu"  on  the  MAIN Menu  and the  enter  key pressed,  the  Variogram
Menu is  activated.   The  programs on the Variogram Menu  are  intended to do the
variogram analyses and  should include the  programming  needed  in the majority  of
cases.    Using the  F5  key  and interfacing  to  GEO-EAS  allows  YOU to  use the
variogram program included in GEO-EAS.   In some cases the GEO-EAS program may
include some  feature not  found  in GEOPACK so  it is advisable to check GEO-EAS  if
you want  a  feature not currently supported in GEOPACK.    Since the data format
requirements of GEOPACK are different from GEO-EAS, an  interface is  necessary and
has been written  into the GEOPACK system.  You must  remember  the  names  you  give
to various  files  while  running GEO-EAS since  it  doesn't  have  a specified naming
convention.    When GEOPACK writes  a  file  which  is  in GEO-EAS format it always has
the extension  EPA (i.e.,  .EPA).

      GEOPACK will save up  to 15  sample  semivariograms  using the following naming
convention:
    \VARij.00
first  saved  sample semivariogram
for variable  i  with j   (i, j = l,..,IVAR)
if variable 1 is  MOIST,  2 is  TEMP
and 3  is SAND,  then if  1=1  and  j=2
VARij.OO is  the first  cross-variogram
between MOIST and  TEMP  saved  by
GEOPACK.
                                        26

-------
    \VARij.01
    CRAM  drive>\VARij.02
                    .  .03
second saved  sample  semivariogram
third saved  sample  semivariogram
    \VARij.14
fifteenth saved  sample  semivariogram
Set Program Parameters:

      GEOPACK provides you with a program for setting the variogram parameters  to
facilitate using  GEOPACK in an  efficient  manner.   This program allows you  to
specify  a  value for  any  of the parameters  used to calculate  the sample semi-
variogram or  cross-semivariogram.  You can specify any or all of these parameters
and once they  are  specified,  the  program will not ask for interactive input during
execution.    For example,  if you  want to  run the program five times and are
planning to  change only  one parameter each time,  you would give a "?" in the
parameter   field for  that  parameter.   During  execution you  will be asked  to
provide  only  that  value.    In this manner,   you will only have to answer the
necessary questions during  execution.
Spatial Correlation Function
Parameter Settings File
Data File Name: (FNAME)
Enable Graphics Options (i.e., -G) (Y/N )
Save All Sample Variogram Functions (Y/N )
Data Cutoff Value: (CUTOFF)
Number of Lagged Distances
Distance Between Lagged Distances:
Width of Lagged Interval:
Width of Angle Class:
Number of Directions:
Angle for Direction Number-1
Angle for Direction Number-2
Angle for Direction Number-3
Angle for Direction Number-4
t up 4- down Home PgUp PC
( NLAG )
( BLAG )
(WLAG)
( DALF )
( NDIR )

;Dn End F2=Save

TEST. DAT
Y
Y
100.0
12
1
0.0
90.0
1
0.0
30.0
60.0
90.0
Esc=Exit
                                        27

-------
Variogram Calculation  (Sample):

      This program calculates the sample semivariogram or cross-semivariogram for
a  spatially-dependent,  two-dimensional  random function.

     There are  two  ways that the results  from the sample semivariogram  program
can be saved.   The easiest method is to specify in the Parameter Settings  file  to
save all  sample semivariograms  (i.e.  , Place  a  "Y"  in  the  appropriate   field)  .
When this  is  done,   every time  the sample   semivariogram program is run,   the
results will  be stored on the temporary subdirectory.  Since 15 files may be saved
at any time,  you will most  likely  never  run the program more than 15 times  so
saving every file will not produce  a problem.   If  you should  create 15 files and
wish to  save  additional  sample  semivariograms,  you can delete  files that  you  do
not expect to use.

     An alternative method  for  saving the output from the semivariogram  program
is to  save  the output from  only those runs that produce  "reasonable" results.
This can be done by selecting OPTION KEY:   "S"  at the  end  of  the Sample Variogram
Calculation program.   The OPTION KEYS  appear  at the  lower  right hand corner  of
the screen during execution of a program.   The  "S"  option  only appears  at  (or
near)  the end of the program  and generally after a  graphic  plot of  the  sample
semivariogram has been produced.


Fit Model Automatically To  Sample Variogram:

      This program  allows you  to fit  a  mathematical model  (e.g.,  a  gaussian,
exponential,  linear,  spherical or power model)  to a  sample  semivariogram.   The
fitting  method uses   the nonlinear least  squares  minimization  technique  of
Marquardt (1963).
                    Sample  Autocorrelation  Functions  Saved
  MOIST / MOIST
  MOIST  / TEMP
  MOIST  / SAND
  TEMP  / TEMP
  TEMP  / SAND
  SAND  / SAND
00  01  02  03  04
00  01

00
   Fl: view   Enter:  select
           D: delete
Home   End   Pg Up
Pg Dn
                                       28

-------
     To   specify   a sample semivariogram  (first  step  in running  this program),
a screen is produced which shows all the saved variograms for each random  function
and all  the cross-variograms  for each  pair of random  functions.   This  screen
displays  the random function name,  such as:
         MOIST  /  MOIST
         MOIST  /  TEMP
for a direct  semivariogram  for MOIST
for a cross-semivariogram for  MOIST and
TEMP
followed by a  series  of  numbers,  such as
         MOIST  /  MOIST
         MOIST  /  TEMP
00  01  02  03
00  03
The numbers indicate the  00th  (or  Olth,  etc.)  saved sample semivariogram for the
specified random function(s).  Given in Section 2.3 is a description  of  the  naming
convention used by GEOPACK. The numbers (i.e. ,  00  or  01,  etc.)  are  the  extension
of the  sample  semivariogram file  name (i.e., VARij.OO).

     Typing RETURN selects the particular sample  semivariogram file for use by
the fitting program.    Typing  Fl  gives some information about  the  sample semi-
variogram file that the cursor  is  highlighting.   This  help  information is  supplied
to help you determine  which file you  want  to use when you have a number of files
from which  to choose.     The output from  the  fitting  program is stored on the
temporary drive  (or subdirectory)  using a  naming convention similar  to the  sample
semivariograms.  After a sample semivariogram has  been modeled (and the results
saved  --   SEE SECTION ABOVE  ON SAVING  THE SAMPLE SEMIVARIOGRAM) the  model
coefficients are saved  in a file
    \mVARij.00



    \mVARij.01

    \mVARij.02
first saved  semivariogram  model
coefficient  file.  This  file corresponds to
\VARij .00

second file  saved

third file  saved
    \mVARij.14
fifteenth file saved where  the  "m"  is used to
designate the MODEL  semivariogram  file.
                                        29

-------
Save Model(s). Fit  Model  Manually:

      This  program  displays  the model  semivariogram  coefficients  that have been
modeled  using the automatic  fitting program and  allows  the user  to  select the
"best" model  to  be  saved  in  a file for use by  the  kriging  program.   The program
allows you to save,  view and manually model the semivariogram.  Each operation is
described below.

      Saving  a Semivariogram.   The semivariogram model can  be  saved  to the file
which  then  becomes  available to  the Ordinary   [Co] kriging  and Disjunctive
 [Co]kriging programs.   This  must  be  done  prior to kriging,  otherwise the kriging
program  will not have  a  variogram  model to use  in the calculations.    If the
semivariogram models  were  fitted  to  the  sample  semivariogram   using  the
"least-square" method (i.e.,  automatic fitting procedure), then the semivariogram
model should be  cross-validated  prior to kriging.  To  do  this,  use  the  Cross
Validation option on the Ordinary  [Co]Kriging menu.  To save the a semivariogram:
type "S",  "RETURN"  and  give  the line number of the model you want  saved.    If no
semivariogram model coefficients are saved,  the  kriging program  will  not work
properly and  an  error  message will be printed  out stating  that no semivariogram
model coefficients were found.

      Viewing a  Semivariogram.    This  option  allows you to  view  a  saved
semivariogram.   You  might  want to do this  if  you  have saved many semivariograms
and are  trying to determine  the  one  you want to use for  kriging.

      To view the a semivariogram:  TYPE  "V",  "RETURN" and give the line number of
the model you want plotted.   If  you  want  to view several semivariogram then give
the line numbers  using the format: "1,2,3" for lines 1  to 3.   Note,  you can only
view 3 or less models on one  graph.

      Manually Modeling a  Semivariogram.   This option allows  you to  alter the
coefficients  of  a saved semivariogram.  You  must use the automatic fitting program
prior to manually modeling a semivariogram since this  gives  you  a  good starting
point for the coefficients.   If several  model files  are  available,  you can see the
sample semivariogram from which  the  model  was  derived by typing the Fl  key.

      If you  choose to manually model the semivariogram, you MUST place the model
coefficients  into the variogram  model  file manually,   as  well  (unless  you decide
to use the  automatically fitted  values  listed  on the screen).

      To manually  model a  semivariogram:   TYPE "M",   "RETURN",   change  the
coefficients shown at the bottom of  the screen, and  type  "ESC"  to  have the model
and sample  semivariograms plotted.
                                        30

-------
Edit Variogram Model File:

      This program allows the user  to directly edit  the file which  stores  the
models for the  semivariogram.    If a model  is  fitted to a  sample  semivariogram
manually,  then the model coefficients must be manually entered into the variogram
model file using  this program.
Variogram Editor: File: G:\TEST.VAR
. . _
-U
— Variables 	 Model 	 Nugget 	 Sili-JMuggec 	 Kange 	 ^ —
MOIST / MOIST : Exponential o
MOIST / TEMP : Exponential o
MOIST / SAND None 0
TEMP / MOIST Exponential o
TEMP / TEMP Exponential 2.001
TEMP / SAND None 0
SAND / TEMP None 0
SAND / SAND None 0
Esc: Exit Fl: Help F2 : Save <- t -» Home
1.371 8.001 00
-2.999 9.994 00
0 0 00
-2.999 9.994 00
15.003 4.513 00
0 0 00
0 0 00
0 0 00
End Page Up Page Down
     If  no  semivariogram model  coefficients are saved, the  kriging  program will
not work properly  and an error message  will be  printed out  stating that  no
semivariogram model coefficients were found.  To correct this error,  use the Save
Model(s), Fit Model Manually program to save the semivariogram model coefficients
for the  random  functions  of  interest.
                                        31

-------
2.5.1.    Description of the  Kriging  Menu.
  Return To Main Menu
                          Thu  Jun 29,  1989 07:47:45 am
     GEOSTATISTICAL PROGRAMS:  Main  Menu
           Ordinary  [Co]Kriging  Menu
   Return To Main  Menu
   Set Program  Parameters
   Cross  Validation
   Ordinary  [Co]Kriging
   Line Contour  Diagram
   Block  Contour  Diagram
   Statistics Menu
   Variogram Menu
   Disjunctive  [Co]Kriging  Menu
Esc: quit
Fl: help
F3: Util
F4 : view
F5: user menu
*
Home
End
      If  the  highlighted  key "0" is typed or  the  light  bar placed over the menu
option "Ordinary [Co] Kriging Menu" on the MAIN Menu and the enter  key  pressed, the
Kriging  Menu is activated giving  access to  several programs for  calculating
estimates of the selected random function(s)  in space.  You can use the F5 key to
interface  to  GEO-EAS  (or other  programs)  which allows  you to use  the ordinary
kriging program for one  random  function included in GEO-EAS.
Set Program  Parameters:

      Information on  using  the setting of the  program  parameters  for kriging is
similar  to that given  in Section  2.5.1.  An example of the screen generated by the
Set Program Parameters program  is  given below.    More information  about the
parameters can be found  by  starting  this  program and using the Fl  key.
                                        32

-------
Kriging Parameter Settings File
Est LOG in File
(Y/N) :
Maximum Search Radius:
Contour File:
(Y/N)
Enable Graphic Option:
ESTIMATE GRID: (P/B)
No. of Points
No. of Points
No. OF VARIABLES
Estimate Var
Auxiliary Var
Auxiliary Var
Auxiliary Var
Auxiliary Var
- X
- Y

n
#i
#2
#3
#4
N
20.00
Y
Y
P
30
30
?
1
2
3
?
?
BLOCK KRIGING DISCRETIZATION
No. of Cells -
No. of Cells -
«-left right->
X
Y
t up 4-dc
3
3
>wn Home
Anisotropy Ratio:
Anisotropy Angle:
Number Contour Levels:
Save Intermed. Results
ORIGIN
Width Betw. Pts - X
Width Betw. Pts - Y
No. NEAREST NEIGHBORS
Estimate Var #1
Auxiliary Var #1
Auxiliary Var #2
Auxiliary Var #3
Auxiliary Var #4
BLK COVAR. (0 default)
Width of Cell - X
Width of Cell - Y
PgUp PgDn End F2 =
1.0
0.0
4
N
0.0,0.0
?
?

10
10
10
10
10
0.0
?
?
Sav Esc=Exit
Cross-validation:

      GEOPACK provides  you with a program  that  allows  you to cross-validate the
model for the spatial correlation  structure.  The cross-validation technique used
is a  semi-quantitative  technique based on kriging and is used  to  evaluate whether
the covariance  function (or  variogram)  is  appropriate  for  the experimental data.
The method  involves estimating the  value  of the random function of interest at
every known  sampling  location but excluding  the  known  value  from the  estimation
process  (the known  value is excluded because kriging is  an exact interpolator).
Using all  the  pairs of actual and  estimated values,  various quantities  can be
calculated and  used as an  indication  of  the  "quality"   of the model  spatial
correlation  function.
Ordinary Kriging and  Cokriging:

      The program  GEOKRIG  is  used for kriging and  cokriging.   GEOPACK allows  up
to  10 variables to be  used  for cokriging although it is advisable  to use 4  or
less random functions in practice.   If more  than  5  random  functions are used,  the
default  settings program cannot be  used  while running the kriging program,  since
it  only allows  you to input  1 primary and 4  auxiliary random functions.  A  large
static array of 10000 words is partitioned for storage of  input data,  the kriging
matrix,  etc.  so there is a limit  to  the  size of problem that  can be executed with
this program.
                                        33

-------
      GEOPACK allows you to create estimated values in one of two manners,  either
on a grid system or by reading a file of  x and y coordinates.   This  latter  method
is useful if  irregularly  spaced estimates are required.

      Estimates  on a Grid.   The first  method is to create  a grid system  with
estimated values and estimation variances  at each of the nodal  points.   To  do
this, you must  supply  the  following  information:

      1.  The  origin of the grid system  (usually  the origin for
         the  collected data is  used).
      2.  The  number  of  columns  of  estimates in the X direction
      3.  The  distance  between  columns  in the X direction
      4.  The  number  of  rows  of  estimates in the Y direction
      5.  The  distance  between  rows  in the Y direction

For this case,  the total  number of  estimates is  (Number of columns)-(Number  of
rows).

      Read  Locations From a File.   An  alternative method  for determining the
placement for calculating  estimates is to create a file which  contains  a list  of
the desired X and  Y coordinates for  the  placement of the estimates.  When  using
this option,  GEOPACK  will prompt  you  for the file name  and use  the file  to
determine the placement of estimates.


Line Contour  Diagram:

      A contouring package is provided with GEOPACK so that the output  files  from
the kriging program can be graphically  illustrated.   The  output file from the
kriging program is in the  temporary storage directory with a name:  KPLOT.TMP. The
contouring program  reads this  file  and produces an  output  file  KPLOT1.TMP  which
can be used by the graphics program to  draw  a  contour plot on any device  for which
a device driver  is  supplied.


Block Contour Diagram:

      The pixel  contour  program is  similar to the line contouring program
described above  except that  the estimation grid  is  presented on the screen and
each cell is filled with a  color and/or pattern in correspondence to the  estimated
value of  the  particular cell.    This provides  another method  for viewing the
results of the data and gives  a better feel  of the coarseness  of  the  grid system.
The pixel diagram  which  results from using this program can be  printed by  using
the GRAPHICS.COM utility supplied with MS-DOS and the PRINT-SCREEN key.
                                        34

-------
2.5.J.   Description of the  Disjunctive  Kriging  Menu.
                          Thu  Jun  29,  1989 07:47:45 am
  Return To Main Menu
     GEOSTATISTICAL PROGRAMS: Main  Menu
         Disjunctive  [Co]Kriging  Menu
   Return To Main Menu
   Set Program  Parameters
   Hermite  Coefficients      (DK Step  1)
   Estimates,   [Co] Kriging    (DK Step  2)
   Line  Contour  Diagram
   Block  Contour  Diagram
   Statistics Menu
   Variogram Menu
   Ordinary  [Co]Kriging  Menu
Esc: quit
Fl: help
F3: Util
F4 : view
F5: user menu
t
Home
End
      If  the  highlighted  key "D" is typed or the  light  bar  placed  over the menu
option "Disjunctive Kriging Menu" on the MAIN Menu and the enter key pressed,  the
Disjunctive Kriging  Menu is activated giving  access to several  programs  for
calculating estimates and  conditional probabilities of the selected  random func-
tion (s)  in space.
Set Program Parameters:

      Information on the setting of program parameters  for disjunctive kriging can
be  found in Sections 2.5.h  and  2.5.1  (i.e., variograms and  kriging).    One
difference between  this  program and the previous programs is  that  there  are two
screens of parameters that can  be  set  for  disjunctive  kriging.   To toggle to the
next screen type Ins.


Calculate Hermite  Coefficients:

      The first step in disjunctive kriging is the  definition of a transform
variable.    The  transform variable is assumed  to be hi-variate normally  dis-
tributed.   The  program calculates  the transformed values of the data set which are
                                        35

-------
used in the  disjunctive  [co]kriging  program.   For each random function that will
be used for  kriging,  a transformed  data  set  must be calculated  (i.e. ,  run this
program for  each data set  selected).   As a  check  on the transform process, a
comparison between the mean and variance  of  the data  and  the transformation is
given and should be  approximately  the  same.
Calculate Estimates:

      The program  GEODK is used to disjunctive kriging and cokriging.    The
disjunctive kriging program allows  up  to 10 variables to  be  used for cokriging.
Up to 10 cutoff values  can  be  specified for which the program will calculate the
conditional  probability  that the  estimated value is greater than the cutoff level.
During execution,  the output to  the  screen will use as heading  designators the
letters A through  J.   Typing the  F10  key will bring up a window that gives the
definitions of the  headings in  terms  of the cutoff values and transformed cutoff
values .
Line and Block  Contour  Diagrams:

      Line  and  block contour  diagrams are  also  available from  the Disjunctive
Kriging  Menu.   For more  information concerning  these  programs,  see  the  description
under the Kriging  Menu,  above.
2.5.k.   Modifying and Adding  Menus.

       It  is  possible to tailor GEOPACK to an individual's  needs  by adding  programs
or menus  to  the  USER menu (accessed through the  F5  key).    If  the program to be
added  to  the  system  requires  the use of the data  set  or  any of the output files
produced by GEOPACK,  read the help information under the heading  SYSTEM OPERATION
on the main  HELP menu.    The information  concerning the files,  their  contents,
formats,   naming  conventions,  etc.  is given by this help sequence.

       The following  steps are used to add  a  program  or menu to  GEOPACK.   Several
examples  are  included in  GEOPACK and can be found by  typing the *.MEN files. It
is usually  easier  to start  with copy of an existing *.MEN  file  (or  portion
thereof)   as  a  template and  make  the  required modification(s) to the template. To
add a single executable  program is very  easy, since only a few lines must be added
to an  existing *.MEN file.   To  add an additional  menu  requires  creation of  a
complete  *.MEN file, and  therefore is  somewhat more  difficult.    If a new menu is
to be created, it is generally a  good idea  to place the  *.MEN and  executable files
is a new  subdirectory.
                                        36

-------
STEP 1: Determining  Whether  a Menu or a Program Will be  Added:

      The  first  step is to determine whether a menu  or a program is to be  added
to GEOPACK.   If  it appears that the  application  to be  added may  require  several
programs or batch files, it is better to create a new menu containing  these  items.
An example of creating a menu is the GEOEAS.MEN file contained in the  subdirectory
%GEODIR%\USER\GEOEAS, where  %GEODIR% is used  by GEOPACK to  designate the  root
directory  of  the geostatistical  programs.  The  reason for creating a menu  rather
than specifying a program is that to run the  program GEOEAS requires rewriting the
GEOPACK data file in  a  format that can be  read by  GEOEAS. This  requires  an
interfacing program  to  create a  new file in the  proper format.

      Next on this menu are  the commands  for  starting  the GEOEAS programs.  This
is followed by a program which takes the output of the kriging program (in GEOEAS)
and rewrites the output in a  format that GEOPACK can  read.  This  enables the user
to utilize the  GEOPACK routines  on the GEOEAS output.    Since each of  these
commands  concerns  the use  of  GEOEAS  it is  better  to put  them  into  a menu
specifically to  run  the GEOEAS program,  rather than  including all these commands
on the USER menu,  since space is limited to 11 entries on any one menu.

Step 2: Altering the USER.MEN:

      The  second step is to alter USER.MEN to include another  or alter an existing
heading.   Once the program or menu  to be added  to  GEOPACK has  been created  it can
be accessed through  the USER  menu  (i.e.,  F5  key)  by  adding or  modifying an  entry
to the  menu  file.    Entries  to the USER  menu  are  added  by editing the USER.MEN
file located in  the  root  of  the  geostatistical  programs   (i.e., %GEODIR%\USER.MEN).
An example of a  menu entry is:

         Line
           1:        %lE@User  Menu
           2:        *_GEO-EAS
           3:        #message
           4:
           5:        A%GEODIR%\user.men
           6:        +%COMSPEC% /c cd  %GEODIR%\user\geoeas
           7:
           8:        A%GEODIR%\user\geoeas\GEOEAS.MEN

where the  symbols  have  the  following meaning:

         I    in column 1 indicates  that the  next two  hexadecimal numbers
              indicate  the Menu  colors,  where  the  first  and  second numbers
              are  the background and foreground  colors,  respectively,
              and  the colors are given by the color-numbers  given below
               (for example a dark  blue background with light yellow
              lettering is  specified by using  "%1E"  in line  1,  above)


                      0  - Black             8  -  Grey
                      1  - Blue              9  -  Light Blue
                      2  - Green             A - Light Green

                                        37

-------
             3  -  Cyan              B  -  Light  Cyan
             4  -  Red               c  -  Light  Red
             5  -  Magenta           D  -  Light  Magenta
             6  -  Brown             E  -  Light  Yellow
             7  -   White           F  -  Light  White
      In  any  other  column the "%" is used  for  different purposes.

@    print the  following text as the menu  title

*    print the  following text as menu  selection  heading

#    print following  text  in the information box  at  the top of
      the menu

~     tells the  menu program to hold the menu  on  the  screen as
      long  as possible before  switching screens.  This  helps to
      improve  the visual  appeal of the  menu program but should
     only be used  when  "popping"  (i.e.  ,  adding)  or "pushing"
      (i.e.,  deleting)  a menu and not when running an executable
     program.   It  should be placed prior  to the  menu to be added
     or  deleted.

+     indicates  that the  remainder of the line is  to  be executed.
     THE MENU  PROGRAM WILL ONLY RUN EXECUTABLE FILES AND NOT
     BATCH FILES OR BATCH  COMMANDS.   To run a BATCH
     command the following method must  be  used:

     + %COMSPEC% /c  els       - this  will clear the screen
     +%COMSPEC%  /c  echo  ... - this will echo .  .  .
     + %COMSPEC% /c cd  c:\ -  this will  call  the directory c:\

     The name of the  file  to be executed must be  either in the
     current directory,  include  a full  path specification or
      the file must be in a directory on the DOS path.   It is
     generally  a good idea to include  the  full path  of
     every program whether you expect  the  program to be in the
     current directory or  not,  since it may not  be obvious what
     the current directory is.   You can always  change the current
     directory  using  "%COMSPEC%  /c CD\
-------
               from  the  right hand side of the  screen  to the left hand
               side;    which  is done to improve  visual  appeal).    (Push the
               menu  off  the  screen).

               the underscore before the  "G"  on line 2 is used to designate
               the highlighted character.   This  character will activate
               the menu  option.   If there are  two  (or  more)  menu options
               with  the  same highlighted character,  the  cursor bar will
               move  to the next  one,  each time  the highlighted character is
               entered.
STEP 3: If Another MENU  Is  Being Added:

       It  is  often  best  to place the new program  into  a  separate  subdirectory.  An
example of this  is  the  GEO-EAS subdirectory which stores the  executable  programs
for the GEO-EAS  geostatistical software system.   Once the subdirectory  is  named
and created,  the newly  developed program (s)  and auxiliary files  should be  copied
into it.   A  new  data  file containing the entries for  the newly created menu must
be created using the  methods  described  above.   It is  probably best  to  start with
an existing menu command file  (e.g.  , USER. MEN,  GEOEAS.MEN  or equivalent) to  act
as a template,  changing only  those  lines  that are needed.   This will provide a
guide  for setting up  the  command file.

     In the  USER. MEN  file,  the call to the new menu is  made  as  follows   (assuming
that a new subdirectory called "NEWDIR"  has been  created with  the following  path:
%GEODIR%\USER\NEWDIR  and that the command file  for the  new menu  is called
"NEWMEN.MEN") :

     #place  a  descriptive message for the menu  here
     *_NEWMEN  Menu

     A%GEODIR%\user.men
     +%COMSPEC%  /c  cd  %GEODIR%\user\newdir

     A%GEODIR%\user\newdir\NEWMEN.MEN

     The  commands  necessary to run one  (or more)  programs  from  the new  menu  are
as follows  (it  is  assumed that the new  program  is  called  NEWPRG) :

     #Quit Menu
     *_Quit
     *Run  _NEWPRG
     #place a  descriptive  message for what  the  program does here
     +%GEODIR%\user\newdir\NEWPRG

The above  instructions should be adequate for modifying GEOPACK to include  the
new menu and program (s) .
                                        39

-------
2.5.1.   Trouble Shooting.

      If  an  error should  occur while  attempting to  add a program or menu  to
GEOPACK try one of the  remedies  listed  below.

1. The newly  created USER menu  entry does  not show up on the  USER menu.    This
might occur if the wrong menu data file was altered.   The proper  file to  alter
is %GEODIR%USER.MEN.   Also  ,  only 11 entries can be  placed on one menu.    If you
wish to have more than  11  entries  you  will  have to  create an additional menu  or
move some of the programs  to  other  menus, if they  exist.

2. The newly created menu  does not  run when selected from the USER  menu. This  is
probably due to an  improper  path-file  name  combination.  Make sure  that the path
and file names are correct.   If  an error message is printed out by the program,
it is often  "lost"  because the  menu reappears on the screen too quickly.    The
error message  (or other  information) can be seen by  typing  the  F4 key.

3. The newly created menu  does not have any entries  on it and  an  error message:
Abort Menu:  Menu Data Not  Found  is printed to  the screen. This  is  caused because
the menu program cannot  find  the newly  created menu  data file.    This may be due
to changing directories  before the new menu is placed on the screen.  You  should
check to make sure  that  the commands in  the User Menu  and  the  new  menu are
appropriate  for what you are  trying to  do.

4. To get back to the USER MENU (located at the right-hand  side of the screen,
you have to type  two  "quits".     This is  caused because  you have popped  an
additional USER.MEN  on to  the screen and the newly created menu only deletes one
of the USER.MEN when you quit the menu.   To delete two USER.MEN place two  V  in
the newly create menu file each  on  a separate  line.
2.6. GEOPACK  Enhancements

2.6.a.   Running GEO-EAS  Geostatistical  Software.

      The GEO-EAS  geostatistical software can be  executed from within GEOPACK.
The GEO-EAS programs are accessed by typing the F5 key  and selecting  the  GEO-EAS
option.    Since the data set  required by GEO-EAS  is  formatted differently  than
GEOPACK,  there is an interface supplied to create a data set in. GEO-EAS  format.
This data set  will  have  the  file extension ".EPA".   There is also an  interface
program to  take the output  from the GEOEAS  kriging program  and  convert it  to
GEOPACK format.   This enables  results  to be printed  and  plotted using  GEOPACK
graphics  capabilities.


2.6.b.   Description of GEOPACK  Utility  Programs.

      GEOPACK  contains a number  of utility programs which  can  be  used separately
from the system to aid in alteration  of  the system or  for use  in a  "command-line"
mode  (i.e.,  running  the program without using  the menu  system) .     A brief
description of the  function of  each  program and  their  use  follows.
                                       40

-------
Program Name
Description
HELP.EXE       This  program is used to  call  a particular help information
               section  from a  file.   For example,   if  you want to print the
               help  information  concerning "Theory of  Variograms"  the HELP
               program   could be  used as  follows:   C:\HELP VARI.HLP
               THEORY.HLP  where  VARI.HLP is  the help  information desired
               and  THEORY.HLP  is the file  containing VARI.HLP.

GEOPACK.EXE    This  program displays  the  first screen giving the  GEOPACK name.
               This program  also  determines where the  call  to  GEOPACK  is
               made  (i.e.,  the subdirectory where  GEOPACK is typed)  so that after
               leaving  the  program you will be returned  to  that  directory.

MSDOS.EXE      This  program allows you  to  execute a  DOS command while run-
               ning  GEOPACK.

DOSSHELL.EXE  This  program allows you  to leave GEOPACK  in a  DOS  shell  and
               temporarily  suspend execution of GEOPACK. When  finished  ex-
               ecuting  DOS  commands  you reenter GEOPACK  by typing EXIT.

MENU2.EXE      This  is  the  menu  program.

EVOKE.EXE      This  program is used to execute a  program when it  is desired
               that  system parameters  be passed to the program.   (Note:
               system parameters are set by the user  and are  difficult  to
               "hard" program).   This  program  is  particularly useful  when
               altering  (i.e.  ,  adding  additional  programs) GEOPACK  since  it
               enables  you  to pass system  parameters  to  other programs
               without  having  to know  their value.  For  example,  to run  a
               "C "   program and  passing  the  location  of the  temporary
               directory,  the border color and the text  color use:

               EVOKE  $(TEMPDIR) $ (BORD) $ (TEXT)

               EVOKE.EXE  substitutes the "true" values for  $(TEMPDIR),
               $(BORD)  and  $(TEXT)  in  the  spawning process, i.e., the above
               command  is  equivalent to  g:\  1 11  provided that
               the  system  is setup so  that the  temporary directory is g:\,
               and  the  border  and text colors are  1 and  11,  respectively.
               The  list of  parameters  EVOKE.EXE can pass follows:

               EDITOR -   Default text  editor
               TEMPDIR   -  Temporary directory
               PRINTER      Printer device
               PLOTTER      Plotter device
               ROWS  -  Maximum number of rows on screen
               COLS  -  Maximum number  of rows  on printed page
               TEXT  -   Text  color
               BORD  -   Border  color
               MODE  -   Screen  mode  (0=mono,  l=color) This only works  in
                         FORTRAN programs

                                        41

-------
               DSTAT  -   Default statistics  package
               LGRAPH  -   Default line graph  package
               GRAPH  - Default  contour,  3-D  graphics pkg
               DBASE  - Default  data base management  system

               Examples:    Try these examples  in  the  root of the geos-
               tatistical  programs

               EVOKE defgraf $ (LGRAPH)   $ (TEMPDIR)  $ (BORD)  $ (TEXT)
               EVOKE defgraf $(GRAPH) $ (TEMPDIR)  $ (BORD)  $ (TEXT)
               EVOKE $(DBASE)  EVOKE  DOSSHELL $(BORD)  $(TEXT)
               EVOKE SDATA $ (TEMPDIR) $ (BORD)  $ (TEXT)
               EVOKE $(EDITOR)
               EVOKE edit  $(EDITOR)     $(BORD)  $(TEXT)

SDATA.EXE      This program  allows you  to declare a data file active by
               copying it  to the temporary directory (e.g. , RAM drive) and
               creating  the DEFAULT  file which  contains  the data file name
               and origination  path.   To use type:  SDATA from root of the
               geostatistical programs.

PACK.EXE       This program allows you to  take  the files on the temporary
               subdirectory (e.g.  ,  RAM  drive)  and store them in an archive
               file.   These files may be retrieved at  a later time using
               one Of  the  programs:   EXTRACT.EXE,  SDATA.EXE  or UNPACK.EXE.

               To use  type:  PACK from root of the geostatistical  programs.

UNPACK.EXE     This program allows you to  take  an archived file and remove
               all the stored  files  back to  the temporary storage drive
               (i.e.,  RAM  disk).

               To  use  type:  UNPACK from root of the  geostatistical
               programs.

EXTRACT.EXE    This program allows you to  take  one or  more stored files out
               of an archived  file and place  them on the temporary storage
               drive  (i.e.,  RAM  drive).    You  must extract  the data file
               and the DEFAULT  file  before attempting  to  run GEOPACK.

               To  use  type:  EXTRACT from root of the  geostatistical
               programs.

EDIT.EXE       This program along with EVOKE allows you  to  call up the
               default text editor  for editing  the data  base.   See examples
               for using EVOKE  for use instructions.

LISTDIR.EXE    This program allows  you to  list  the contents  of any direc-
               tory on a floppy or hard  drive.   The  program also allows you
               to view a file.

               To use  type:    C:\LISTDIR    
-------
               colors-.  where the colors  are  given by the color-numbers
               given below:
                      - Black
                      - Blue
                      - Green
                      - Cyan
                      - Red
                      - Magenta
                      - Brown
                      - White
8 -  Grey
9  - Light Blue
10  - Light Green
11  - Light Cyan
12  - Light Red
13  - Light Magenta
14  - Light Yellow
15  - Light White
VIEW.EXE       This  program allows  you to view  the contents of  a file. To
               use  this program type  the  following: C:\VIEW  -
               + file name> where the  border  and text colors
               are given  in above.   Colors are  either decimal as  described
               above  or hexadecimal  using a format  such  as:  OxlE.

VED.EXE        This  program allows you  to  manually edit the  semi-variogram
               coefficient  file (see the  variogram menu and associated
               help  screens for more information)  if  it  exists.

AUTO.EXE       This  program operates with the menu system and executes  the
               necessary   commands to  create  the hard-coded and user-
               definable  menus.

MGEOEAS.EXE    This  is an example of  a user-written interface to  couple  the
               GEO-EAS geostatistical  software to  the  GEOPACK system.   This
               program modifies that GEOPACK  data  file so the GEO-EAS  can
               read  its contents.

MGEOCNVT.EXE  This  is an example  of  a user-written interface to couple  the
               GEO-EAS geostatistical  software to  the  GEOPACK system.   This
               program modifies the GEO-EAS output  file  so that GEOPACK  can
               use  the results for further analysis.

SHOWDIR.EXE    This  program allows you to see a  diagram  of all  the sub-
               directories  that branch off the  current  (or specified)
               subdirectory.   To use type: SHOWDIR  .

DEFLT.EXE      This  program allows you  to alter  the system settings for  the
               parameters  listed under EVOKE.

               To use this  program type:  DEFLT  SYSETUP  SYSDEFLT.
                                        43

-------
                                  SECTION 3

                                  EXAMPLES

3.1 VIRUS DECAY  RATES IN TUCSON GROUND WATER

Description  of Data

      Water  samples were  obtained  from 71 continuously pumping public water supply
wells in the Tucson Basin.   The water samples were  taken  to the  laboratory and
viruses were added.    The samples were  incubated  at the in situ ground-water
temperature.   At  predetermined  times,  subsamples were withdrawn  and  assayed to
determine the number of virus particles remaining in  the water.   The  decay rate
was calculated as the slope of the line  drawn through  the  data using  linear
regression and has  the units of Iog10 (virus particles)/day.

Geostatistical  Calculations

      The data  set was  configured in the following way  according to GEOPACK
requirements  and  is included in the directory %GEODIR%\USER\DATA under  the name,
VIRUS.DAT.

  Ground Water Data for  the  Tucson Basin
2 71
1 2
TEMP
500.00
3 4
DECAY
5




(f5.0,4F10.3)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
8.
14.
14.
5.
5
6.
7.
5.
17.
2.
-0
5.
6.
3.
14.
11.
14.
14.
-4.
-5
-1.
5.
9.
.498
.129
.330
.280
.079
.084
.291
.883
.951
.263
.151
.280
.285
.671
.934
.716
.129
.934
.374
.983
.961
.280
.906
7
6
8
8
8
7.
6
5
6
8
9
7
6
7
5
6
4.
4.
11.
12.
10.
6.
4.
.458
.453
.263
.665
.464
.056'
.855
.045
.654
.866
.671
.458
.654
.660
.447
.654
.240
.442
.682
.889
.475
.252
.643
20.
20.
20.
20.
20
23
23
23
23
24
24
24.
24.
24
24.
24.
24
24.
24.
24.
26.
26.
26.
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.50000
.50000
.50000
.50000
.50000
.50000
.50000
.50000
.50000
.50000
.50000
.00000
.00000
.00000
-o.
-0.
-0.
-1.
_ ]_
-0.
-0.
-0
-0.
-0.
-0
-0.
-0.
-0.
-0.
-0.
-0.
-0.
-0
-0
-0
-0.
-0
.82800
,94600
.92400
.16400
.06700
.79700
.98900
.89200
.88900
.69500
.85500
.73400
.76900
.80200
.69600
.67600
.96300
.81000
.80500
.81400
.79600
.69900
.76000
                                       44

-------
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
8
8
10
15
15
16
12
16
-5
-5
3
3
5
10
8
7
8
9
11
11
11
15
15
15
15
16
16
5
4
6.
6.
9
10
13.
17.
1.
-1
-1
1.
3
3
3.
4.
2.
3
5.
5.
2.
.096
.096
.107
.135
.738
.543
.520
.744
.581
.581
.269
.269
.481
.710
.096
.492
.297
.101
.917
.515
.314
.537
.135
.939
.537
.744
.543
.079
.878
.285
.285
.302
.509
.526
.951
.056
.961
.357
.861
.068
.470
.470
.274
.665
.872
.682
.883
.062
5
5
5
3
4
3
2
0
11
11
6
7
5
3
3
4
4
4
-1
3
3
3
2
1
2
0
1
5
4
3
4
1
2
3
2
7
8
7
2
2
5
6
5
1
1
0
2
4
.246
.849
.246
.436
.240
.235
.430
.620
.280
.481
.855
.257
.447
.034
.838
.643
.643
.240
.592
.034
.838
.034
.430
.425
.430
.419
.023
.447
.643
.637
.240
.626
.430
.235
.430
.458
.665
.660
.833
.631
.447
.051
.246
.626
.023
.821
.028
.039
26
26
26
26
26
26
26
26
26
26
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
28
28
28
28
28
28
28
28
30
30
30
30
30
30
30
30
30
30
30
30
30
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.25000
.25000
.25000
.25000
.25000
.25000
.25000
.25000
.50000
.50000
.50000
.50000
.50000
.50000
.50000
.50000
.50000
.50000
.50000
.50000
.50000
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0.
-0
-0.
-0.
-0.
-0
-0
-0.
-0.
-0.
-0.
-0.
-0
-0.
-0.
-0.
-0.
-0.
.64000
.94800
.48800
.59900
.57600
.94500
.75400
.67600
.79700
.54400
.39800
.37100
.81600
.31700
.67800
.65800
.76500
.78700
.73400
.68900
.73400
.62900
.68400
.33900
.77200
.53400
.23600
.60400
.46800
.65600
.71300
.58400
.79400
.66400
.72700
.19400
.67000
.66000
.53800
.33400
.63800
.55200
.15100
.50500
.43900
.53300
.62900
.41400
      Several  steps  are required  for  conducting the geostatistical  analysis  and
are described below.   The  data set was created  following  the  format  instructions
given in Section 2.4.a.   All user inputs are  shown  in  Boldface type.

                                        45

-------
Step 1:  Select  a  Data  Set

       To use  GEOPACK a data  set  must be available.   Therefore  the  first  step in
using  this program  is  to  create  a data set  (which is assumed to  have  been done)
and then start  GEOPACK  and select  it.

 1) Select Data Set Utilities Menu  at Main  Menu.
 2) Select Select a Data Set at Data Set Utilities  Menu.
 3)   To select the data  set,   either use  the cursor keys to  find the  entry
VIRUS.DAT or  type the  name of the  file on  the keyboard.   If a  number of files
exist in this directory, it is possible to move the cursor to the  first (or next)
file beginning with a specific letter by typing ALT-letter  (i.e., hold the ALT key
and type the desired letter).  Note:  the  Select  Data program automatically looks
in the  "data" directory.  This directory  is  specified  in the SYSDEFLT file which
can be  altered  using  the  Change  System Default Settings program  (i.e.  ,  type F3
then C, its the second line on the screen  that follows).   Once the file VIRUS.DAT
has been selected, GEOPACK reads in  the data  file, and  outputs  to the screen the
cutoff  level and  the number of available  data points  for the  variables.   You are
returned to the Main Menu.
Step 2:  Check  Distribution  of  Data

      The  next step in the analysis is to  determine if  the  data are normally
distributed.    Although the  kriging technique does not required the data  to  have
a normal distribution  (and  is  a  best  linear  unbiased  estimator regardless of the
distribution of the data)  the  estimator  is optimal  whenever  the data  is normally
distributed.

 4) Select Statistics  Menu.
 5) Select Kolomogorov Test of Distribution.
 6) Select the appropriate data set, in this  case,  B,  decay.    The  calculations
are output to  the screen.   At this  time,  you may  type J to  jump  ahead.   The
results  of the  Kolomogorov-Smirnov  Test for Distribution  are output  to  the  screen.
In this case,  we  accept the null hypothesis  at the 0.05 level  and thus  there  is
some  evidence  that  the data  are normally  distributed.  If the  data were not
normally distributed,   it  would be  possible to  transform  the  data  into a new  data
set with  a normal  distribution.    A  logarithmic  transformation  is a  common
technique for  transforming  data  into  a normal  distribution.  You  are  returned  to
the Statistics Menu.
Step 3:  Determining  the  Variogram

      Next  the sample spatial correlation structure  (i.e., the  sample semi-
variogram)  must be determined.  Without this  functional  relationship  between  the
distance and the variance it  is  not  possible  to use the kriging techniques.

 7) Select  Variogram Menu.
 8) Select  Variogram Calculation  (Sample).
 9) Select  type  of  correlation function,   in this  case,  A,  semivariogram.

                                       46

-------
10)  Select the  principal function, in  this case,  B,  decay.
11)  You  will  now be asked  several questions  about how you want  the  calculations
performed,  unless  you  have  set up the  default  settings  file  in such a  manner  as
to avoid this.  When responding to the  questions,  you  may either  type  in a number
or push  Enter to select the default value  which is shown on the  lower  left-hand
side of  the screen.
12) After all of these  questions  have  been  answered,  the  variogram is  calculated.
The progress  of the calculations  is output  to  the  screen  as  Percent Completion  on
the  lower right-hand side.
13)  Both a table of values  and a plot  of the  sample variogram  are printed to the
screen.   For  this data  set,  the output appears as shown in Figure  3.1.1  (Note: the
following  figures  use  the device drivers supplied  with GEOPACK.    Publication
quality  figures can be created using  an appropriate graphics  package and  the USERS
MENU (F5).
14)  You  are  now given  several  options.   Select Options  to add  labels  to your
graph.    Upon  Exiting this  menu,  you  are returned  to  the  Variogram Menu.
15)  Select Automatic Model Fit.
16)    You will  again  be  asked  several questions, then allowed to  choose the
autocorrelation function you wish to  model.
17)  Select a  model type,  in this case, D, spherical.   The input parameters are
printed  to the  screen,  as well as the  input data.   If you wish  to delete any  of
the data, type  D,  Enter,  the Observation number,  and  Enter.   You need  to repeat
this step  for each line  you  want to  delete.    In this  case,   we delete  lines 8
through  12.
18)  Several screens will be printed,  detailing the results of  the model  variogram
calculations .   The model  fit to  the  data will then be printed  to  the  screen.  For
this example,   the  output  is as shown  in Figure  3.1.2.
19)   Once again, you are given  the opportunity to add labels,  etc. to the graph,
20)   Upon Exiting the  screen, you are  asked if you want  to rerun  the  program.  If
No,  you  are returned to the Variogram Menu.
21)   Select Manual Model  Fit,  Select Model(s).
22)   You now  select the desired file  by hitting Enter.
23)   A summary  of  the  models  that you  have run  on that  data  file is then shown.
Type S to save  the correlation model  that you want to krig.
24)  You  are  returned to the Variogram Menu.


Step 4:  Create  a Grid of Estimates Using  Ordinary Kriging

      The final  step  in this example  is to obtain a  grid of estimates  using the
data set and  the model of  the  spatial correlation structure.  An  estimate can  be
made at  any point  in the field of interest.   In general,   estimates made far from
any  data points will  have a higher  estimation variance (i.e.  ,  there  is  less
confidence that the estimate  is  close to the true value)  compared to  estimates
made close to data points.   If an estimate  is  made at  the same  coordinates as one
of the data  points in  the data  set,   the estimated value will  exactly  equal the
data value at that point  and  the estimation variance will be zero. This behavior
indicates that  kriging  is  an exact interpolator.


25)   Select  Ordinary  Kriging.   You will  be  asked several  questions unless you
have modified the  Default  Settings file to avoid  this.

                                        47

-------
26)    Several screens showing  the output from  the  kriging calculations  will  be
output to the screen.   Once again,  at the Kriging Estimates for Decay screen, you
are given  the opportunity  to  Jump ahead and  bypass  seeing the output  of  this
program.
27)  When the  program is  finished,  you are returned  to  the  Ordinary  Kriging Menu,
where you select Line Contour  Plot.
28)  You  are given the choice  of plotting the  estimated value or the variance.
In this case,  we choose  A,  and the following contour plot  is shown on the  screen:
29)  You  are given  the option to add labels  to  the  plot as before.
30)  After Exiting  this  program, you are returned to the  Ordinary Kriging Menu.
31)  This time,  we select Block  Contour  Plot.
32)  Again we choose A to plot the estimated  values,  and the output looks like the
following:
                                       48

-------
e = .0-
DATE: 09/01/1989


SEMIVARIOGRAM

No. OF
LAG COUPLES
1
2
3
4
5
6
7
8
9
10
11
12
11
125
232
230
249
241
216
193
160
175
137
146
DISTANCE
.41
1.23
2.38
3.54
4.65
5.83
6.95
8.17
9.38
10.50
11.66
12.77
CALCULATION 56 » ± 90.0'
TIME: 09:06 am
SEMIVARIOGRAM DRIFT
DECAY DECAY
.0243 -.1444
.0292 -.0649
.0366 -.1125
.0415 -.1355
.0426 -.1603
.0437 -.1633
.0424 -.1621
.0379 -.1557
.0342 -.1586
. 0406 - .1680
.0574 - .2497
.0426 -.1555
                                                                                      A
             Strike a key when ready .  .
                                                                             X R|0|H|
                                         Virus Data
                8,845
                8,83
                8.825
                                                                         B
                                        Distance (km)
Figure  3.1.1.   Example  output  from Variogram Calculation (Sample)  program  for the
data set,  VIRUS.DAT.   in A  and B, respectively,  are  the  tabular values and  a  plot
of  the semivariogram function.

                                         49

-------
           8,85
          8,845
           0,84
          8,035
         18,8
          8,825
           0,82
          8,815
           8,01
                                      Virus Data
                                     Distance (km)
                                                                    18
Figure  3.1.2.    Example  output  from the Automatic Model Fit  program.   The  solid
line  is an spherical model  that was  fitted to  the sample  semivariogram.
                                          50

-------
SUMMARY OF SPATIAL CORRELATION
Line No. Sill -
No. Pts DIR MODEL NUGGET NUGGET
1 10 .00 EXPON .0165 .0240
[ 2.33] [ 3.55]
2 10 .00 SPHER .0207 .0197
[ 5.34] [ 4.98]
3 10 .00 POWER .0000 0311
[ .00] [ .16]
Esc: Exit Fl : Help S: Save V: View
MODELS
RANGE
1.2286
[ 2.00]
3.8252
[ 3.52]
.1297
[ .19]
M: Model

SSQ
.00008
.00007
.00016
(Manual Mode)
                                                                              A
   = Variables
   TEMP  /  TEMP
   TEMP / DECAY
   DECAY / TEMP
   DECAY /  DECAY
 Variogram  Editor:  File:  G:\VIRUS.VAR
 Model    '    Nugget == Si 11-Nugget  = Range =
     None            0             0             0
     None            0             0             0
     None            0             0             0
Spherical        0.021          0.02         3.825
                                                                     00
                                                                              B
ESC:  Exit
            Fl:  Help
                       F2 :  Save
                                             Home
                              End | _ Page Up | | Page Do;
Figure  3.1.3.   Example output  from Manual Model  Fit,
A) and  the Edit Variogram  Model File  (in B) .
                                                      Select  Model (s)  program  (in
                                     51

-------
                                 Virus Decay Kate
                                   Distance (km)
Figure 3.1.4.   Contour diagram  of  the estimated  value of the  decay rate using  the
Ordinary  [Co]Kriging  and Line  Contour Diagram  programs  is  illustrated.
                                         52

-------

                                          HI•<•)«>
-------
3.2.  SALINITY  IN A SOUTHWEST ARIZONA FIELD.

Description of  Data

      This  data set contains  two random functions,  the electrical  conductivity
 (EC),  and the  sodium adsorption  ratio (SAR).    The  data was collected by Al-
Sanabani  (1982) at  101  random locations  on a 1m  by  1m grid  system in a  10-ha
field  (soil type:  Typic  Haplargid).   The  1-2  kg  samples were placed  in  a  beaker
with distilled  water  and  a  saturated  paste  was  made.   The  solution  was  extracted
from the paste  and  the EC of  the  solution  measured.   A  more  detailed description
of the data and techniques used to measure  the  EC and  SAR  is  given in Al-Sanabani
 (1982) .   This data  set is included"  in  the  %GEODIR%\USER\DATA directory  under  the
name, ECSAR.DAT.
Step 1:  Check  Distribution of Data

       The  hypothesis  that the data  were  normally or lognormally distributed
was  tested using  the  Kolomogorov-Smirnov  (KS)  test for  goodness-of-fit  (Sokal
and Rohlf, 1981; Rao  et  al.,  1979).   The KS  test statistic was calculated
and compared to  the  critical  value at the  0.1 probability level.   This  level
was chosen to  reduce the  probability of the type  II  error,  namely  the probability
of incorrectly selecting  the  null  hypothesis  (Rao  et  al.  ,  1979).

       For  a  sample size  greater than 30 the  critical value at the a= 0.1
probability  level  can  be obtained from the asymptotic  expansion  (Rohlf  and
Sokal, 1981; Rao et  al.,  1979):

     K Scrit  =  0.805/(7n)

       The  results  of  the KS test  indicate  that the  original  and log-
transformed  data are  from lognormal and  normal  distributions respectively,
at the 0.1 probability level.  Therefore a  lognormal  transformation will  cause  a
data set  to have a normal distribution.  The KS test  statistic calculated  from the
data sets were 0.197 and  0.065 for the original (EC) and transformed  (InEC) data,
respectively.   The associated critical value  is  0.080  and was calculated by Eq.
1  for n  =  101.    Since  the data  are lognormally distributed, they  will be
transformed  to  a data set with a  normal  distribution by taking the  logarithm of
each datum.  This  new data set will be used  for  the geostatistical analyses.

Step 2:  Transform  the  Data Set

       The data set can be transformed  by using the Modify an Existing  Data Set
option on the  Data Set Utilities  Menu.   The  steps necessary to use  this program
to add a  random function called In [EC]  by  taking the  logarithm  of the  EC data  are:

1.   The  allowable  transformations  can be viewed by  typing Fl  (Help)  before
selecting the variable  to modify.   Place the lightbar  over  the  data  set  (i.e., EC)
to be  transformed.  Type  Enter.
2. Move  the  lightbar over one of  the  Unused markings  and type  Enter.    A window
will pop  up in which the name for  the new data set can  be  entered.   In  this  case,
the new  data set will  be  called  In[EC].

                                        54

-------
3. Type  in  the  transform equation,  which for this example  is:

                     y=log(x);              (natural  logarithm)

and type a F4 to  start  the  transformation process.   Once completed both the  EC and
In [EC] data sets  will  exist in the  data file.
4. To save the transformation  results,  type  F2  and  either  0 to  overwrite  the  data
file or  B to  save a backup copy of  the old data file.   If you decide you  do not
want to  save the  new data  set,  the  type ESCAPE and you will be asked  if  you  want
to leave the  program without making any changes.
5. To exit, type  ESCAPE.


Step  3:  Determining the Variogram

      As with all  kriging methods   some  measure  of the spatial  correlation
structure is necessary.   For disjunctive kriging the appropriate  function  is the
auto correlation  which  requires  a second-order stationary hypothesis  (i.e. ,  the
variance must be  finite).

      An  easy method for determining  the autocorrelation  from  the semi-variogram
is to use the following relationship

    p(h)  =  1  -r(h)A<«)

which is done automatically when using the  disjunctive kriging  program.  Using
GEOPACK  to determine  the sample semivariogram by following the sequence  given in
Example   1.  Next  a model variogram is  applied  to  the  sample semivariogram.   For
this  example, the automatic fitting procedure  is used  and yields

    •y(h)  =  0.27  +  0.66 [1.5h/145  -  0.5 (h/145)3] ;    0 145

and is shown  in  Figure  3.2.1.

      This variogram model was used in  the  cross validation procedure listed on
the Ordinary  [Co]Kriging Menu which attempts to fit  both the sample semi-variogram
and the  reduced  mean and variance simultaneously.   An  acceptable  model  variogram
is given when  the  reduced mean and variance are approximately 0.0  and  1.0,
respectively,  subject  to a  reasonable  fit  to the  original sample  semivariogram.
Using the Cross Validation procedure on the Ordinary [Co]Kriging Menu,  the reduced
mean  and variance  of  the  above model  can  be determined  and are, respectively,
 .0198 and 1.224.

      The cross-validated model is somewhat  different  from the one listed above
because now the  fitting procedure also tries  to fit  the  reduced  mean and variance.
For this case the semi-variogram model  is

    f(h) = 0.36 +  0.52 [1.5h/153 - 0.5 (h/153)3] ;    0 153

                                        55

-------
      For the model given above,  the reduced mean and variance,  respectively, were
0.0241  and  1.062.    One problem in using the  reduced mean  and variance for
determining the adequacy of semivariograms is that there is no  independent method
for determining how  close  to  zero and unity the reduced mean and variance should
be and  therefore  using this  technique  to fit models to a  variogram is somewhat
subjective.

      Viewing the two variograms using  the Fit Models Manually option on the
Variogram Menu,  it appears that the neither of the models produces an  adequate fit
for both the  sample  semivariogram and reduced mean and variance. For this reason,
an alternative model is chosen that falls  in between  the cases listed  above.  This
model is

    •y(h) = 0.30 +  0.60 [1.5h/160  - 0.5(h/160)3] ;    0 < h < 1 6 0               [5]

    •y(h) = 0.90 = ?(«);                                h >  160

and has a reduced mean and variance,  respectively,  of 0.0203 and 1.19.


Step 4:  Calculate the Hermite Coefficients

      The next step  in the  disjunctive kriging process  is  to find the  coefficients
which define  the  Hermite  transform function  (See Theory  of Geostatistics listing
on the GEOPACK User's Manual,   Type Fl) .   To  obtain the  Hermite coefficients  it
is necessary  to select the Hermite Coefficients listing on  the Disjunctive Kriging
Menu.   There  are  a number of  questions that must  be  answered during the  execution
of this program.   These  questions will  appear only  if the  Set  Program  Parameter
program was not run  prior  to  this selection.


Step 5: Create  a  Grid of Estimates Using  Disjunctive Kriging

      A total of  2500 estimates of EC and  In(EC)  (using each method) were obtained
on a 5 by 5  meter grid system superimposed over the field sampled by Al-Sanabani
 (1982). Along with each  estimate a value  of  the kriging  variance was calculated
and for the  disjunctive  kriging method a  value of the conditional probability.
Bower and Wilcox  (1965) gives  4.0  dS/m  as  a critical value of  EC for many plants
therefore  this value  (or  the In(4.0 dS/m)  )  was used for  the  critical value (i.e.   ,
Zcut)  in subsequent  analyses.   For all cases  a maximum of 5  nearest  neighbors
within a radius equal to the  range of the variogram were  used  in the estimation
process.   Next,   the results  were  contoured to show the spatial  distribution of
the electrical  conductivity and  illustrated in Figure 3.2.2.


Step 6: Create  a  Grid of Conditional Probability  Estimates

      The disjunctive kriging estimator  can  also be used to determine the
conditional  probability that the unknown value is greater than a specified cutoff
value.    An  example of  this is  given  in Fig.   3.2.3 where the  conditional
probability that  the ln[EC]  is greater than  ln(4.0),  respectively  is contoured.

                                        56

-------
From this  figure  one can see  that  the zones of  high  probability coincide with
areas of high  In [EC] .   Although the shapes  of  the high  probability zones  are
similar to the shapes of the In [EC]  contours  in Fig.  3.2.2, two points with  the
same  estimated value  of In [EC] do  not  always  produce  the  same  conditional
probability.   For more information see Yates  et.  al.  (1986)
                                        Salinity
                                       Distance (n)
Figure  3.2.1.    The semivariogram function for the  natural  logarithm of  the
electrical conductivity using the Variogram  Calculation  (Sample) program  and  the
data set, ECSAR.DAT.
                                        57

-------
                      level 1
                      Level 2
                      Level 3
                      ' eve! 4
                       ?vel 5
                       evel 6
                       evel 7
                       evel 8
-8.1894 <=  jiiiiilJIK 8.6179
 8.6179 <=       < 8.826
  8.826 <=
  1.049 <=
  1.229 <=
  1.347 <=
  1.491 <=
  1.752 <=
  1.849
  1.752
< 2.636
Figure 3.2.2.     Contour  diagram  of  the estimated  value of   the  electrical
conductivity  using  the  Disjunctive   [Co]Kriging  and  Block Contour  Diagram  programs.
                                          58

-------
Figure 3.2.3.   Contour diagram of the  conditional probability  that  the  estimated
value of the electrical  conductivity is greater than the natural logarithm of  4
dS/m  (i.e.,  1.38 log(dS/m})  using the Disjunctive  [Co]Kriging and  Block  Contour
Diagram programs.    This  figure shows the actual  screen  position of the  contour
levels window when  displayed.
                                       59

-------
3.3.  SURFACE  MOISTURE,  TEMPERATURE  AND  SOIL TEXTURE.

Description of  Data

      This data set contains three random functions, the  surface moisture content,
 (9) ,  the  surface  soil  temperature (T)  and the sand content of the  surface  1  cm.
The moisture  content and  sand  content data  were  collected  at  71  random locations
and the temperature data  at 120 random locations on a 1m  by  1m grid system in a
1  ha  field at the Campbell Agricultural  Center  of the University  of  Arizona. A
more  detailed description of  the data and techniques used to measure  the  EC and
SAR is  given in Yates  and Warrick  (1988) .    This data set  is  included  in  the
Available Data  Sets option  (Type:  F5,  followed by A)  and  is named  C912.DAT.


Step  1:  Determining  the Variogram

      As  with  all kriging methods  some measure of the spatial correlation
structure is necessary.   For cokriging a spatial  correlation function is required
for each  random function as well a  cross-correlation function for each pair of
random functions to be included in the analysis.   Since ordinary kriging is to be
performed it  is only  necessary that the random  functions  be  intrinsic (i.e.,  it
is  no longer  required that  the variance  be  finite) .

      GEOPACK is used  to  determine the sample  semivariogram  by following  the
sequence  given  in  Example 1,    Next a model  variogram  is  applied  to  the  sample
semivariogram.   For  this  example,  the automatic  fitting  procedure is used  and
yields the  following  exponential  models

    •y(h) = 0.00  + 34.2  (1 - exp [-h/8 . 78] ) ;            for  surface  moisture

    •y(h) = 0.00  - 18.4  (1 - exp [-h/10 .8] ) ;            for  cross-semivariogram

    •y(h) = 2.86  + 15.6  (1 - exp [-h/4 . 71] ) ;            for  surface  temperature

for h > 0  and are  shown in Figures  3.3.1  and 3.3.2.  The  overall  reduced mean and
variance for  this set of variograms  is -0.0091 and  0.92, respectively.

      This variogram model  was used in the cross validation procedure listed on
the Ordinary  [Co] Kriging Menu  which  attempts  to fit both the sample  semi -variogram
and the  reduced mean  and variance  simultaneously.   Since only one model  can be
cross-validated  at a  time  the  following method was  used.  First  the  semivariogram
for the  surface moisture was  cross-validated.    This  was followed  by cross-
validating the  surface  temperature  model.    Assuming that  these  semi -variograms
were  not  affected  by  the  cross-validation of the  cross-semivariogram,  they were
fixed during  the  cross-validation of the cross-semivariogram.  Using  this  Cross
Validation procedure the  following  variogram models
         =  0.00 + 34.0 (1 - exp [-h/8. 23] );            for surface  moisture

    Y(h) =  0.00 - 18.3 (1 - exp [-h/10 .3] );            for  cross-semivariogram

    y(h) =  4.19 + 14.3 (1 - exp [-h/5 . 12] ) ;            for surface  temperature

                                        60

-------
for h > 0 were calculated.   The  overall  reduced mean and variance for this set of
variograms is -0.0068 and 0.88,  respectively.   Plots of the model variograms and
the  sample  variograms are  shown  in the Figures.   Viewing  the cross-validated
variogram models using  the Save Model(s).  Fit Models Manually option on the
Variogram Menu indicates that either set of variogram models  would be adequate
representations  of the spatial  correlation structure.

Step 2: Create a  Grid of Estimates  Using CoKriging

      A total of  2500 estimates of  the  surface moisture content were obtained on
a 2 by 2  meter grid  system  superimposed over the CAC field site using cokriging.
 (the auxiliary variable  was the surface temperature). Along with each estimate
a value of the kriging variance was  calculated.   Shown  in Fig.  3.3.3 is a block
contour diagram  of  the cokriging estimates and  definition of  the color/pattern
used in the diagram.   For this  example a maximum of  5 nearest neighbors of the
moisture  and  temperature within a  radius of  20  m from the estimation  site was
used in  the estimation  process.    Shown  in Figure  3.3.4 is the associated
estimation variance  for  the  soil moisture using  the  cokriging  technique.
                                       61

-------
                32
                28
             £  21
             g1
             1  28
             *  16
                12
                 22
              en  IA
              .2  12
              is
              1  18
                                       Soil tloisture
                            12
                             12
                                       Soil Itnperature
                                         Distance (n)
                                                          0        0 1
                                                    -0	2	
                                                                         B
Figure  3.3.1.   The  semivariogram function for the  surface  moisture content  (in  A)
and  the surface  soil  temperature  (in B)  using  the Variogram Calculation  (Sample)
program and the  data  see,  C912.DAT.
                                           62

-------
                             Soil Boisture and lenperature
                                    Distance (n)
Figure  3.3.2.   The  cross-semivariogram function  for the  surface moisture content
and the surface soil  temperature using the Variogram Calculation  (Sample) program
and the data set,  C912.DAT.
                                         63

-------
                         Level  1
                         Level  2
                         Level  3
                         Level  4
                         Level  5
                         Level  6
                         Level  7
                         Level  8
1.4471
5.9127 <
7.1459 <
8.1546 <
9.3162 <
18.872 <
13.079 <
16.831
<=
<=
<  5.912?
<  7.1459
<  8.1546
<  9.3162
<  18.872
<  13.079
<  16.831
<  24.674
Figure  3.3.3.     Contour  diagram of  the  estimated value  of  the electrical
conductivity using  the  Ordinary  [Co]Kriging  and  Block Contour Diagram programs.
                                            64

-------
                       Level  1
                       Level  2
                       Level  3
                       Level  4
                       Level  5
                       Level  6
                       Level  7
                       Level  8
    0 <=
13.282 <=
17.942 <=
21.077 <=
23.692 <=
25.912 <=
 28.34 <
34.687 <
   < 13.282
;!:!:; < 17.942
   < 21.877
   < 23.692
   < 25.912
   < 28.34
   < 34.687
   < 61.784
Figure  3.3.4.     Contour  diagram   of the estimation  variance  for  electrical
conductivity using  the  Ordinary  [Co]Kriging  and  Block  Contour  Diagram  programs.
                                          65

-------
                        REFERENCES  AND SUGGESTED READING

Abramowitz,  M.  and A.  Stegun, Handbook  of Mathematical Functions,  Dover
         Publications,  Inc.,  New York, 1965.

Al-Sanabani, M.,  Spatial Variability  of Salinity and  Sodium Adsorption
         Ratio  in a Typic Haplargid  Soil,  M.S.  Thesis, The  Univ.  of
         Arizona,  Tucson,  1982.

Baafi, E.R.,  Program User's  Manual  for  Basic  Geostatistics System, Unpublished
         Report,  Dept.  of Mining and Gee.  Eng.,  Univ.  of Arizona,  50 pp. ,  1982.

Bell,  K.R,  B.J.  Blanchard,  T.J.Schmugge  and M.W. Witczak,  Analysis of
         Surface  Moisture Variations  Within Large-Field Sites,  Water
         Resour.  Res.  16:796-810, 1980.

Bhattacharyya, G.K.  and R.A.   Johnson,  Statistical  Concepts  and  Methods,
         John Wiley,  New York, 639 pp.,  1977.

Bower, C.A.  and  L.V.  Wilcox,   Soluble  Salts,  In: Methods  of Soil Analysis
         Part 2,  C.A.  Black,  Ed., Monograph No. 9,  Am. Soc.  Agronomy,
         Madison,  WI,   pp. 933-951,  1965.

Burgess,  T.M.  and R.  Webster, Optimal Interpolation and  Isarithmic Map-
         ping of  Soil  Properties. I.  The Semivariogram and  Punctual
         Kriging,  J.  Soil Sci. 31:315-331,  1980a.

Burgess,  T.M.  and R.  Webster, Optimal Interpolation and  Isarithmic Map-
         ping of  Soil  Properties. II.  Block Kriging. , J.  Soil  Sci.
         31:333-341,  1980b.

Burgess,  T.M.,  R. Webster and A.B.  McBratney,  Optimal  Interpolation and
         Isarithmic Mapping  of Soil  Properties.  IV.    Sampling Strategy. ,
         J.  Soil  Sci.  31:643-659, 1981.

Burrough,  P.A.,  Multiscale  Sources of Spatial  Variability  In  Soil.  I.
         The Application of  Fractal  Concepts  to Nested Levels  of  Soil
         Variation,  J.  of Soil Sci.,  34:577-597,  1983a.

Burrough, P.A.  ,  Multiscale  Sources of Spatial  Variability  In  Soil.  II.
         A  Non-Brownian Fractal  Model  and  its Applications  in  Soil
         Survey,  J.  of Soil  Sci., 34:599-620,  1983b.

Carr,  J.R.,  D.E. Myers and  C.E.  Glass,  Co-kriging  - A Computer  Program,
         Comp.  and Gee.  , 1984.

Clark, I.,   Practical  Geostatistics, Applied  Sci.  Publ.  LTD, Essex,
         England,  129  pp.,  1979.

David, M,  Geostatistical Ore  Reserve  Estimation,  Elsevier,   New  York,  1977.
                                         66

-------
Davis, B.M.,  Indicator Kriging as  Applied  to an Alluvial  Gold
        Deposit,  Geostatistics for Natural Resources  Characterization,
        Part  I,  G.  Verly et al.  Eds.,  pp.  337-348,  1984.

Dunn,  M.R.,  A Simple  Sufficient Condition for a Variogram Model to Yield  Positive
        Variances  Under Restrictions Math.  Gee.,  15:553-564,  1983.

Dunn,  M.R.,  A Simple  Sufficient Condition for a Variogram Model to Yield  Positive
        Variances  Under Restrictions:  Reply Math. Gee.,  16:443-434,  1984.

Gajem, Y.M.,  A.W.  Warrick and D.E.  Myers,  Spatial Dependence  of  Physical
        Properties  of a Typic Torrifluvent  Soil,  Soil Sci.  Soc. Am.  J.
        45:709-715,  1983.

Hatfield,  J.L.,  J.P Millard and  R.G.  Goettelman,  Variability  of  Surface
        Temperature  in Agricultural Fields  of Central  California,
        Photo.  Eng.  and Remote Sensing,  48: 1319-1325,  1982.

Hatfield,  J.L.,  M.  Vauclin, S.R.  Vieira  and R. Bernard,  Surface  Tempera-
        ture  Variability Patterns  with  in  Irrigated Fields, Agric.
        Water Manage.   8:429-437,   1984.

Journel, A.G.  and Ch.J.  Huijbregts, Mining Geostatistics,  Academic
        Press,  New York,  1978.

Journel, A.G.,  The Place  of  Non-parametric Geostatistics,  Geostatistics  for
        Natural  Resources  Characterization, Part  I,  G. Verly  et al.  Eds.,
        pp.  307-335,  1984.

Journel, A.G.,  Fundamentals of Geostatistics  in Five  Lessons,  Department of
        Applied  Earth Sciences Report,  Stanford University  Stanford,  CA, 1988.

Kim, Y.C.,   D.E.  Myers and H.P. Knudsen,  Advanced Geostatistics  in  Ore
        Reserve  Estimation and Mine Planning  (Practitioner's  Guide),
        Report  to  the U.S.  Energy  Research and Development  Administra-
        tion,  Subcontract No.  76-003-E,  Phase II,  1977.

Lemmer, I.C.,  Estimating Local Recoverable  Reserves Via  Indicator
        Kriging,   Geostatistics  for Natural  Resources Characterization,
        Part  I,  G.  Verly et al.  Eds.,  pp.  349-364,  1984.

Marechal,  A.,  The  Practice of  Transfer Functions: Numerical Methods  and
        Their  Application,   Proceedings, NATO A.S.I.  "Geostat  75",  D.
        Reidel  Publishing Co., Dordrecht,  Netherlands,  1976.

Marquardt,   D.W.,  An Algorithm  for Least-squares Estimation  of Non-linear
        Parameters,  J. Soc. Ind.  Appl.  Math,  11:431-441,  1963.

Matheron,  G., A Simple Substitute  for  Conditional Expectation:  The
        Disjunctive  Kriging,  Proceedings, NATO  A.S.I.  "Geostat 75",  D.
        Reidel  Publishing Co., Dordrecht,  Netherlands,  1976.
                                         67

-------
McBratney, A.B.  and R. Webster,  How Many Observations  are  Needed for
        Regional  Estimation of Soil  Properties?,  Soil  Sci.,
        135:177-183,  1983.

Morkoc, F.,  J.W.  Biggar,  D.R.  Nielsen  and D.E. Rolston,  Analysis of
        Surface  Soil Water  Content  and Surface Soil  Temperature using
        State-Space Approach,  Submitted to Soil Sci. Soc.  Am.  J.,  1985.

Myers,  D.E.,   Matrix Formulation  of  Co-Kriging, Math. Gee.,  14:249-257,
        1982.

Myers ,  D.E.,   A Simple Sufficient  Condition for a Variogram Model to
        Yield Positive Variances  Under Restrictions: Comment,
        Math.  Gee.  16:431-432, 1984a.

Myers,  D.E.,   Co-kriging  -  New Developments,  G. Verly et  al.  (Eds.),
        Geostatistics  for Natural  Resources  Characterization,  Part  1,
        pp.  205-305,  1984b.

Rae, P.V., P.S.C.  Rae, J.M.  Davidson  and L.C. Hammond, Use of  Goodness-
        of-fit Tests for Characterizing the  Spatial Variability of  Soil
        Properties,  Soil  Sci.  Soc. Am.  J.  43:274-278,   1979.

Rendu,  J.M.,   An  Introduction  to  Geostatistical Methods of  Mineral
        Evaluation,  S. African Inst.  of Mining and Metallurgy,  Northern
        Cape  Printers LTD,   Kimberly,  South Africa,  84pp.,  1978.

Rendu,  J.M.,   Disjunctive  Kriging: A  Simplified Theory,  Math. Gee.
        12:306-321,  1980.

Rohlf,  F.J.  and  R.R Sokal,   Statistical Tables 2nd  Ed., W.H.  Freeman and
        co.,   New York, 219  pp.,  1981.

Russo,  D., A Geostatistical  Approach  to the  Trickle  Irrigation  Design
        in Heterogeneous  Soil.  1. Theory.  ,  Water Resour. Res.,
        19:632-642,  1983.

Russo ,  D., A Geostatistical  Approach  to the  Trickle  Irrigation  Design
        in Heterogeneous  Soil.  2.  A Field Test.,  Water Resour.  Res.
        20:543-552,  1984a.

Russo ,  D., Design of an  Optimal  Sampling Network for  Estimating the
        Variogram,  Soil  Sci.  Soc. Am.  J.  48:708-716, 1984b.

Russo ,  D., Statistical Analysis  of  Crop Yield-Soil  Water Relationships
        In Heterogeneous  Soil  Under  Trickle  Irrigation,  Soil Sci.  Soc.
        Am.  J.  48:1402-1410, 1984C.

Russo ,  D., A Geostatistical Approach to Solute Transport in  Hetero-
        geneous  Fields and  Its Applications  to Salinity  Management,
        Water Resour.  Res.   20:1260-1270,  1984d.

-------
Russo, D.  and  E.  Bresler.   Soil  Hydraulic Properties  as Stochastic
         Processes:  I.   An Analysis  of Field Spatial Variability.    Soil  Sci.
         SOC. Am.  J.   45:682-687,  1981.

Sisson,  J.B. and  P.J.  Wierenga,  Spatial  Variability of  Steady-State Infiltration
         Rates  as  a Stochastic Process,  Soil  Sci. Soc.  Am.  J.  45:699-704, 1981.

Sokal, R.R,  and F.J.  Rohlf, Biometry  2nd Ed.,  W.H. Freeman  and Co.,
         New  York,  859  pp.,  1981.

Sullivan,  J.,  Conditional recovery  estimation  through probability
         kriging - Theory and  practice,  Geostatistics  for Natural Resources
         Characterization,  Part  I,  G.  Verly et  al.  Eds.,  pp. 365-384, 1984.

Task Committee  on  Geostatistical  Techniques in Geohydrology  -  American  Society of
         Civil  Engineers,  Review  of  geostatistics in geohydrology 1. Basic
         Concepts,  In  Press,  American  Society of Civil  Engineers,  1990.

Task Committee  on  Geostatistical  Techniques in Geohydrology  -  American  Society of
         Civil  Engineers,  Review  of geostatistics in geohydrology  2.  Applications,
         In Press,   American Society of Civil Engineers,  1990.

ten  Berge,  H.F.M., L.   Stroosnijder,  P.A. Burrough, A.K.  Bregt and
         M.J. de Heus,   Spatial Variability of Physical  Soil Properties
         Influencing  the Temperature of the Soil Surface,  Agric.  Water
         Manag.   6:213-226, 1983.

Thomas,  G.B.,  Calculus and  Analytical Geometry, Addison-Wesley Inc.,
         Reading,  Mass., pp.  1034, 1972.

Vauclin,  M., S.R.  Vieira,  R.  Bernard  and J.L.  Hatfield,  Spatial
         Variability  of Surface Temperature Along two  Transects of  a
         Bare Soil, Water Resour.  Res. 18:1677-1686, 1982.

Vauclin,  M., S.R.  Vieira,  G. Vachaud  and D.R.  Nielsen, The  Use of  Cokriging With
         Limited Field  Soil Observations,  Soil Sci.  Soc. Amer. J. 47:175-184,  1983.

Vieira,   S.R., D.R. Nielsen and J.W.  Biggar,  Spatial Variability of  Field-Measured
         Infiltration  Rate,   Soil Sci.  Soc.  Am.  J.,  45:1040-1048,  1981.

Webster,  R.  and T.M.  Burgess,  Optimal Interpolation and  Isarithmic Map-
         ping of Soil  Properties.  III.   Changing Drift  and  Universal
         Kriging.  J.  Soil Sci.,  31:505-524,  1980.

Warrick,  A.W.,   D.E. Myers and D.R. Nielsen, Geostatistical Methods Applied to Soil
         Science,  To be in Methods of  Soil Analysis, ASA, 1985.


Yates ,  Marylynn V.,  S.R.  Yates,  A.W.  Warrick and C.P.   Gerba,   Modeling  Virus
         Transport  in the  Subsurface  using Geostatistics,  Applied and  Environmental
         Microbiology,  52:479-483,  1986.
                                         69

-------
Yates, S.R.,  A.W.  Warrick and  D.E.  Myers, Disjunctive  Kriging.    I.  Overview of
         Estimation and Conditional  Probability, Water  Resources Research,
         22:615-622,  1986.

Yates, S.R.,  A.W.  Warrick and  D.E.  Myers, Disjunctive  Kriging.
         II.  Examples,  Water  Resources  Research, 22:623-630,  1986.

Yates, S.R.,  Disjunctive  Kriging.  III. Cokriging,  Water Resources Research,
         22:1371-1376,  1986.

Yates, S.R.,  A.W.  Warrick and  D.E.  Myers,  A Disjunctive  Kriging Program  for Two
         Dimensions,  Computers  and  Geosciences,  12:281-313,  1986.

Yates  , S.R.  and  A.W.  Warrick, Estimating Soil Water  Content  using  Cokriging,  Soil
         Science  Society of America  Journal,  51:23-30,  1987.

Yates  , Marylynn  V.  and S.R.   Yates,  A Comparison of  Geostatistical Methods  for
         Predicting Virus  Inactivation  Rates  in Groundwater,  Water Research,
         21:1119-1125,  1987.

Yates, S.R.,  A.W.  Warrick, A.D.  Matthias  and S. Musil,  Spatial Variability  of
         Remotely-Sensed  Surface Temperatures at Field Scale,  Soil  Science Society
         of  America Journal,   52:40-45,  1988.

Yates, S.R.  and  Marylynn V. Yates,  Disjunctive Kriging as an Approach to Decision
         Making,  Soil Science Society of America Journal,  52:1554-1558, 1988.

Yates  , Marylynn  V.  and S.R.   Yates,  Septic tank setback  distances: A way  to
         minimize virus contamination of drinking water,  Ground Water,
         27:202-208,  1988.

Yost, R.S,  G.  Uehara and  R.L.  Fox,  Geostatistical Analysis  of Soil
         Chemical Properties   of Large Land Areas.  I.  Semi-variograms,
         Soil  Sci.  Soc. Am. J.,   46:1028-1032,  1981a.

Yost, R.S,  G.  Uehara and R.L.  Fox,  Geostatistical  Analysis  of Soil
         Chemical Properties   of Large Land Areas.  II.   Kriging,  Soil
         Sci.  Soc.  Am.  J.,  46:1033-1037,  1981b.

Zirschky, J.   Geostatistics   for  Environmental Monitoring  and Survey
         Design.   Environ.  International,  11:515-524, 1985.

Zirschky, J.  and D.J.  Harris.   Geostatistical  Analysis  of  Hazardous
         Waste  Site Data.   J. Environ.  Eng.  112:770-784,  1986.

Zirschky, J.,  G.P.  Keary,  R.O. Gilbert  and E.J. Middlebrooks,  Spatial  Estimation
         of  Hazardous Waste Site  Data,  J.  Environ.   Eng.  111:777-789,  1985.
                                         70
                                              £u.S.GOVERNMENT PRINTING OFFICE: 1992 -  648-003/60057

-------