Modeling Distribution of Point and Nonpoint Sources Pollution Loadings in the Saginaw Bay Watersheds, Michigan - Geospatial Tools for Urban Water Resources - page 100

Table 6.1 Input variables derived by the AVDLBRM interface
Variables Databases
Elevation USGS digital elevation model (DEM) a
Flow direction USGS DEM
Slope USGS DEM
Land use USGS land use database b
Depth of upper soil zone (USZ) USDA STATSGO c
Depth of lower soil zone (LSZ) USDA STATSGO
Available water capacity (%) of USZ USDA STATSGO
Available water capacity of LSZ USDA STATSGO
Permeability of USZ USDA STATSGO
Permeability of LSZ USDA STATSGO
Soil texture USDA STATSGO
Manning's coefficient value Land use, slope, and soil texture
a U. S. Geological Survey National Elevation Dataset (NED) http://seamless.usgs.gov/
b U. S. Geological Survey National Landcover Characterization Dataset (NLCD) 1992, http://
seamless.usgs.gov/
c U. S. Department of Agriculture 1994. http://soils.usda.gov

Table 6.2 Time series meteorological and flow variables
Variables

Databases

Daily precipitation National Weather Service climate databases

Daily air temperature National Weather Service climate databases

Daily solar isolation National Weather Service climate atlas

Daily flows USGS discharge database

interflow, percolation, deep percolation, USZ and LSZ moisture storages, ground-
water storage, and lateral flows between adjacent USZ, LSZ, and groundwater
(Croley and He 2005 ).
The DLBRM hydrology component was calibrated for the period 1950-1964,
applied to the period 1999-2006, and recalibrated for this last period to reproduce
the observed daily flow. Performances (Table 6.3 ) indicate that the model
reproduces the flow of the Saginaw River and AuGres-Rifle Rivers well and
with sufficient robustness for nutrients load assessment. Performances for the
Kawkawlin-Pine and Pigeon-Wiscoggin are less satisfying, probably due to the
very small portion of these watersheds contributing to the flow measured at
the USGS gages (He and DeMarchi 2010 ).

6.4 GIS-Model Interface

10 3 km 2 )
watersheds, development of the input variables for each grid cell from multiple
databases over large watersheds is a challenge. To facilitate the input and output
processing for the DLBRM, an ArcView-DLBRM (AVDLBRM) interface program

Since the DLBRMwas designed for hydrologic modeling of large scale (

>