Creating Area Source Spatial Surrogates
As stated previously, you will need to have the PROJ4-GEOS-PostGIS packages installed and operational on your data base prior to running any spatial aspect of CONCEPT. Also, as stated previously, you will need to have downloaded the shapefiles from EPA’s anonymous FTP site in order to create area source spatial surrogates. The total size of these shapefiles (~6.1 GB) simply prohibits us from packaging these data into single pseudo-installation package at this time.
Please note: For testing purposes, the data files that are included in this release of the area source spatial surrogates code will generate only a population surrogate for the 36 km RPO domain in the United States; therefore, you need really only download the census data from EPA’s anonymous FTP site (i.e., shapefiles named as “us_pophu2k.*”).
Please note: You will need to modify the directory structure in the run_control.txt file that is included with this release to point to the SPATIAL_SURROGATES.TXT and SPATIAL_COVERAGES.TXT files for your particular installation.
SPATIAL_SURROGATES.TXT
This file, which is located in the directory $CONCEPT_HOME/concept_projects/beta/scenario1, controls how spatial surrogates are prepared. It is an ASCII, “:” delimited file that has the following structure:
Field 1 – spatial surrogate code (SSC) (VARCHAR(10))
Field 2 – spatial surrogate formula (text)
The spatial surrogate code is arbitrarily defined by the user. The SSC defined by the user in this ASCII file will be used as the SSC in the XREF.RPO_SR. Therefore, the user must ensure that the SSC is consistently used between, in particular, this ASCII file and the CONCEPT SQL table XREF.RPO_SS.
The spatial surrogate formula is, again, user defined. It tells CONCEPT how to build each spatial surrogate. It takes the following general form:
weight * shapefile.fieldname
* weighttype + weight * shapefile.fieldname
* weighttype + …
where weight is
the weighting fraction to apply (numeric); shapefile.fieldname defines the
field in the shapefile that is to be used to
construct the grid spatial surrogate (text); and weighttype defines whether the fieldname is an AREA type (e.g.,
population is area based), a LENGTH type (e.g., roadways are length based); or
a POINT type (e.g., marine shipping ports are point based) (text). Please note that shapefile must have been
previously read into CONCEPT using the import_spatial_coverages
function of the CONCEPT control script, which means that shapefile must have been defined
in Field 2 of the file
$CONCEPT_HOME/../concept_projects/<project>/<scenario>/spatial_coverages.txt
Example entries in SPATIAL_SURROGATES.TXT include the following:
100:1*us_pophu2k.total_samp*AREA
200:0.75*us_pophu2k.total_samp*AREA+0.25*us_pophu2k.households*AREA
300:0.6*us_pophu2k.total_samp*AREA+0.4*us_roads_2000
For the case of SSC=100, the population based surrogate is created. It uses the population data (TOTAL_SAMP) from the shapefile us_pophu2k to create the gridded area source surrogate. For the case of SSC=200, a composite surrogate based on population (TOTAL_SAMP) and housing count (HOUSEHOLDS), both of which are AREA based values, from the shapefile us_pophu2k is defined with weighting fractions of 0.75 for population and 0.25 for housing applied. For the case of SSC=300, we see where fields from different shapefiles and of different weight types can be used to construct an area source spatial surrogate.
Running the program
As with all CONCEPT functions, running the area source spatial surrogate code is a matter of executing the CONCEPT control script with the correct parameters:
% concept create_spatial_surrogates –n <project> -s <scenario> -d <directory>
The test case that is included in this release ran in
approximately one hour on a high-end computer.
Again, the test case will build a population-based surrogate for the