Data and Model for the Final Report of Toward Next Generation Downscaling for Hydrologic Prediction in the Pacific Northwest
Summary
Data Sources, inputs, parameters, and code for the MACA-VIC project final report. Consists of 3 tasks: 1. Consists of comparing results of the indexing method MTCLIM to estimate incoming short and long wave radiation, to observations, to BSRN station data, and to three Ameriflux towers in the Pacific Northwest. 2. Reports on forcing VIC with downscaled GCM forcings, with variations in two forcing variables: downward shortwave radiation (rad) and specific humidity (qair). For this task we consider the MACA downscaling method. Three cases are reported: a) both variables are downscaled; b) rad is indexed and qair is downscaled; c) rad is downscaled and qair is indexed. 3. Reports on forcing VIC with downscaled GCM forcing, with 5 [...]
Summary
Data Sources, inputs, parameters, and code for the MACA-VIC project final report. Consists of 3 tasks:
1. Consists of comparing results of the indexing method MTCLIM to estimate incoming short and long wave radiation, to observations, to BSRN station data, and to three Ameriflux towers in the Pacific Northwest.
2. Reports on forcing VIC with downscaled GCM forcings, with variations in two forcing variables: downward shortwave radiation (rad) and specific humidity (qair). For this task we consider the MACA downscaling method. Three cases are reported: a) both variables are downscaled; b) rad is indexed and qair is downscaled; c) rad is downscaled and qair is indexed.
3. Reports on forcing VIC with downscaled GCM forcing, with 5 variations in the downscaling methods. For this task, in contrast to task 2, rad and qair are always downscaled. Runs with five downscaling methods reported: a) dBCSD; b) MACA; c) mBCSD; d) NARCCAP; e) sdbcNARCCAP (or bcNARCCAP).
Tasks 2 and 3 share the following characteristics:
1. The downscaled data corresponds to the CCSM3e GCM model run. All downscaling was “trained” with the gridded Livneh et al., 2013 data set, downloaded in July 2012.
2. Only 4 pixels of the domain were selected for the figures in the final report of the project and stored in this repository, even though the model was set up and run for the whole pacific northwest domain. Though we carried out runs for 2 scenarios, here we only present the future SRES A2 scenario results.
3. The version of the code used was VIC version 4.1.2.g, with several improvements which have been incorporated into VIC version 4.1.2.h. The compiled executable is included in this directory and the sourcecode can be found online at http://www.hydro.washington.edu/Lettenmaier/Models/VIC/.
4. The model parameters correspond to those used by Livneh et al., 2013, for the continental US portion of the Pacific Northwest. These runs were carried out at the 1/16 degree resolution (~6 km).
