_BEST_ Download Prism Data
Rachele Weishaar
PRISM is a set of monthly, yearly, and single-event gridded data products of mean temperature and precipitation, max/min temperatures, and dewpoints, primarily for the United States. In-situ point measurements are ingested into the PRISM (Parameter elevation Regression on Independent Slopes Model) statistical mapping system. The PRISM products use a weighted regression scheme to account for complex climate regimes associated with orography, rain shadows, temperature inversions, slope aspect, coastal proximity, and other factors. Climatologies (normals) are available at 30-arcsec (800 meters) and monthly data are available at 2.5 arcmin (4 km) resolution. PRISM is the USDA's official climatological data.
Users requiring data with somewhat less precise spatial climatology, but more suitable for climate monitoring, may wish to check out NOAA's nCLIMGRID and nCLIMDIV datasets, which are based upon GHCN-Daily. Users looking for alternative, high-resolution datasets may wish to consider Daymet (see link in "Related pages" sidebar).
The high-resolution (800m) timeseries data require a fee and are no longer distributed in netCDF. The 30-year, 800m monthly climatologies ("normals") are still available from the PRISM website to download for free.
Mean absolute errors in the PRISM annual precipitation normals have been published regionally (Daly et al. 2008). They range from 5% in the eastern US to 10% in the West. Temperature errors range from 0.5C in the east to 1.0C in the West. Increased terrain complexity and sparser data availability contribute to higher uncertainty in the West. These normals guide the development of other climate datasets by providing the expected spatial patterns of climatic variables under average conditions for the CAI analysis. PRISM daily (1981-present) and monthly (1895-present) time series datasets are developed using this technique, which means that uncertainties in the normals propagate into these products.
It is difficult to estimate the uncertainty of the time series datasets with any degree of certainty, given the wide temporal variations in data density and complexity of the climatic fields (especially precipitation). However, studies have been conducted to estimate the uncertainty in PRISM gridded datasets in situations where high-quality ground-truth data are available. Daly et al. (2017) compared monthly PRISM precipitation grids (AN81m dataset) to a high-density rain gauge network operating in the 1950s at the Coweeta Hydrologic Laboratory, located in the southern Appalachians of western North Carolina. The PRISM grids matched closely (within 5%) with the Coweeta dataset. Interestingly, the PRISM regression prediction interval, which is used to estimate interpolation error, overestimated the actual ground truth error. Strachan and Daly (2017) compared PRISM daily temperature grids (AN81d dataset) to observations of daily temperature maximum (TMAX) and minimum (TMIN) on 16 sites located on steep, open woodland slopes ranging from 1967 to 3111 m in elevation in the Walker Basin, California-Nevada. Individual site mean absolute errors varied from 1.1 to 3.7C with better performance observed during summertime as opposed to winter. There was a consistent cool bias in TMIN for all seasons across all sites, with cool bias in TMAX varying with season. This study highlighted the role that differences in the site characteristics of the PRISM source data in the area (in this case SNOTEL stations located in flat areas surrounded by forest), and those of the comparison stations (primarily steep, open slopes) can play in producing local temperature biases.
Average winter (Dec-Jan-Feb) precipitation in western Colorado based on the 800-m PRISM dataset. Units are mm water equivalent. Locations of popular ski areas are indicated. (credit: NCAR Climate Data Guide, D. Schneider)
Grids aredeveloped using PRISM (Parameter-elevation Regressions on IndependentSlopes Model). PRISM interpolation routines simulate how weatherand climate vary with elevation, and account for coastal effects,temperature inversions, and terrain barriers that can cause rainshadows. Station data are assimilated from many networks acrossthe country. For more information, see the Descriptions of PRISMSpatial Climate Datasets.
Data generated within 30 days of observation havethe status "early". Data generated within 1-6 months of observationmay have the status "provisional" and data older than 6 monthsare marked as "permanent".
These PRISM datasets are available without restrictionon use or distribution. PRISM Climate Group does request that theuser give proper attribution and identify PRISM, where applicable,as the source of the data.
The Prism Virtual Data Room is a simple and secure data room solution that streamlines document and data sharing. Prism offers all users a high standard of performance and security, with technology built to serve the needs of sophisticated financial firms in private equity and venture capital.
df19127ead