Willamette Basin Streams
The datasets below are all compressed into a single ZIP file,
WMETTEBASIN.STREAM.DATASETS.ZIP. Decompression produces the following
four files, each in spreadsheet format (Dbase IV), totaling about 20 MB.
REGDATF.DBF – Data for 151 sites sampled for fish in the Basin. Includes
physiographic and LULC estimates for each site. To build most fish regression models,
we used only the 130 sites having INCLUDE=”YES”, because the other 21 sites
employed a substantially different sampling effort than that of the standard EMAP
protocol. The data set also includes sample data for aquatic invertebrate indicators (EPT
richness and WINOE), but a few of the sites used to develop the invertebrate models are
not included here (contact A. Herlihy for further details; Herlihy.Alan@epa.gov). See
Appendix A for variable definitions.
ALLCHAR4.DBF – Physiographic driving variables for all 4045 stream reaches (2-4
order) in the basin. Each reach corresponds to one arc segment in the MODELSTRMS
coverage. In making model projections, all driving variables are assumed fixed over all
scenarios. Variables are defined in Appendix B.
ALLSCEN4.DBF – Land use/land cover (LULC) driving variables for all 4045 stream
reaches. Estimates are given for all 5 alternative scenarios. Variables are defined in
Appendix C.
PROJALT.DBF – Model projections for all model indicators and all scenarios, on all
stream reaches. Also includes all components of the summer streamflow budget (Sec. 3
below). Model projections do not include uncertainty propagation. Variables are defined
in Appendix D.
ARC coverages – Each coverage is in ARC interchange format (.e00) which has then
been compressed via gzip:
ALLPTCOV – Point coverage of 151 stream sample sites.
SITELIST – Point coverage of the 130 stream sample sites used in developing most
Alternative Futures fish models.
MODELSTRMS – Coverage of 2-4 order model stream reaches, defined as individual arc
segments in the full River Reach 2 stream network for the Basin. Segments greater than 5
km in length were split so that no model reach exceeded 3 km. Segments coded as
ditches, aqueducts or reservoir centerlines were not modeled. Model projection results
and/or driving variable data can be merged with the attribute table of this coverage, via
the SEGID attribute.