This directory contains summaries of 30 arc-second (~800m) snow products generated using the
Variable Infiltration Capacity (VIC) macroscale hydrologic model.
The driving meteorological dataset was generated by downscaling the 1/16th degree data used
by Littell et al. 2011 (Regional Climate and Hydrologic Change in the Northern US Rockies
and Pacific Northwest, http://cses.washington.edu/picea/USFS/pub/). The downscaling to
30 arc-second was achieved using PRISM climate normals for monthly average temperature
and precipitation.
In addition to the increased resolution, VIC was adjusted to take gridcell mean slope and aspect
as inputs to the radiative calculations.
Due to space constraints, only SWE (Snow Water Equivalent, mm) was archived from the current VIC
simulations.
-------
DIRECTORY STRUCTURE:
There are 4 subdirectories, one for each of the following scenarios:
(*NOTE: see Littell et al. for more details on the formulation of these.)
1) pnw_historical:
These are the results based on the historical daily time series of temperature and
precipitation, which runs from 1915 to 2006
2) pnw_comp_A1B_2030-2059:
Results from the run in which the historical time series are perturbed using the mean,
gridcell-resolved, changes in temperature and precipitation for the composite of the
top 10 performing Global Climate Models (GCMs), forced by the A1B emissions scenario.
3) pnw_pcm1_A1B_2030-2059:
Results from the run in which the historical time series are perturbed using the mean,
gridcell-resolved, changes in temperature and precipitation for the Parallel Climate Model,
Version 1 (PCM1), forced by the A1B emissions scenari. Temperature projections from this
model tend to be *cooler* than those from other GCMs.
4) pnw_miroc_3.2_A1B_2030-2059:
Results from the run in which the historical time series are perturbed using the mean,
gridcell-resolved, changes in temperature and precipitation for the Model for
Interdisciplinary Research on Climate, version 3.2 (MIROC_3.2), forced by the A1B
emissions scenario. Temperature projecitons from this model tend to be *warmer* than
those from other GCMs.
As with the historical run, the perturbed future scenarios are all run using 92 years of daily
driving data.
-------
SUBDIRECTORIES:
Each run directory contains three sub-directories:
monthly_summaries -- containing summary files at monthly time steps, as follows:
- swe_monthly_day1: the value of swe on the first day of each month
swe_stats -- containing summary statistics for swe
- max_swe: maximum swe for the given year (mm)
- jd_max_swe: date of maximum swe (day of year)
- jd_10pc_accum date of 10% accumulation (swe > 10% of max swe)
- jd_90pc_melt date of 90% melt (swe < 10% of max swe)
- numdays_10pcaccum-90pcmelt jd_90pc_melt - jd_10pc_accum
-------
FILE TYPES:
- (no suffix): standard VIC summary files: one row per year, one column per gridcell
row 1: latitude of each gridcell
row 2: longitude of each gridcell
row 3-Nyrs: data
- .xyz mean of all years in 3-columned ascii format: lat lon data
- .asc mean of all years in gridded arc-ascii format (described below)
- .png quick diagnostic plot of data
-------
ARC-ASCII FORMAT:
The arc-ascii format is readable by ArcInfo (see ESRI help website for information). Each file
in this dataset has the followng 6-line header* which defines the grid:
ncols 988
nrows 937
xllcenter -124.65
yllcenter 41.2
cellsize 0.008333333
NODATA_value -9999
* NOTE: the ascii-grid header specifies "xllcenter/yllcenter" NOT "xllcorner/yllcorner": the Arc
default is to assume the latter.
-------
DATASET UPDATES:
01 DEC 2011:
A bug was discovered in the model setup for the VIC runs used to generate this dataset.
Specifically, the setting that governs the sub-grid calculations of snow based on
elevation bands ("snowbands") was switched off. This meant that snow calculations were
based on the mean elevation of each grid cell. At the high resolution of this dataset,
this correction is probably not too major, especially in areas with smoothly varying
terrain. However, especially in areas with complex topography, addition of the snow bands
should improve the representativeness of the snow simulations.
As a result, all of the results were recomputed with the number of elevation bands set
to a maximum of 5 (i.e., up to 5 separate sub-grid elevation bands for snow computations).
All data were updated and replaced by Jan 10th, 2012.
-------
CONTACT:
Feel free to send any questions my way, at:
gmauger at uw dot edu
(206)685-0317
Guillaume Mauger
Climate Impacts Group
JISAO / UW Seattle