Meteorology at the land surface affects many processes in the terrestrial biogeochemical system. Measurements of near-surface meteorological conditions are made at many locations, but researchers are often faced with having to perform ecosystem process simulations in areas where no meteorological measurements have been taken. To overcome these limitations, the Daymet model was developed by Dr. Peter Thornton while with the Numerical Terradynamic Simulation Group (NTSG) at the School of Forestry, University of Montana.
Daymet is a collection of algorithms and computer software designed to interpolate and extrapolate from daily meteorological observations to produce gridded estimates of daily weather parameters. Weather parameters generated include daily surfaces of minimum and maximum temperature, precipitation, humidity, and radiation produced on a 1 km x 1 km gridded surface.
The required model inputs include a digital elevation model and observations of maximum temperature, minimum temperature, and precipitation from ground-based meteorological stations. The Daymet method is based on the spatial convolution of a truncated Gaussian weighting filter with the set of station locations. Sensitivity to the typical heterogeneous distribution of stations in complex terrain is accomplished with an iterative station density algorithm (Thornton et al., 1997).
Watch a segment of a NASA Earthdata Webinar: The Daymet data set, ORNL DAAC, and the Daymet Web Site.
Daymet Output Variables
The Daymet output variables included for distribution are minimum and maximum temperature, precipitation, water vapor pressure, shortwave radiation, and snow water equivalent. In the Daymet algorithm, spatially and temporally explicit empirical analyses of the relationships of temperature and precipitation to elevation are performed. In addition, a daily precipitation occurrence algorithm is introduced, as a precursor to the prediction of daily precipitation amount. Surfaces of humidity (water vapor pressure) are generated as a function of the predicted daily minimum temperature and the predicted daily average daylight temperature. Daily surfaces of incident solar radiation are generated as a function of Sun-slope geometry and interpolated diurnal temperature range. Snowpack, quantified as snow water equivalent, is estimated as part of the Daymet processing in order to reduce biases in shortwave radiation estimates related to multiple reflections between the surface and atmosphere that are especially important when the surface is covered by snow (Thornton et al., 2000). The Daymet dataset includes estimated SWE as an output variable since this quantity may be of interest for research applications in addition to its primary intended use as a component of the Daymet shortwave radiation algorithm.
Please see Literature for a more detailed explanation of the Daymet algorithm.
The Daymet model requires spatially referenced ground observations of daily maximum and minimum temperature and precipitation. For Daymet V3, surface observations were available from a single source; the Global Historical Climatology Network (GHCN)-Daily dataset distributed by the National Centers for Environmental Information (NCEI) (Menne et al., 2012). Surface observations occuring within Mexico were acquired by GHCN-D through the Servicio Meteorológico Nacional. Through an agreement with GHCN-D, these data were processed throught the same QA/QC measures as all GHCN-Daily data, thereby ensuring a more robust data provenance. One file per year for each year is assembled and input into the Daymet model algorithm. Additional inputs for the Daymet algorithm are a land/water mask, a digital elevation model (DEM), and derived horizon files at the desired spatial resolution and projection of output surfaces.
The Daymet algorithm manages the large number of input data and large spatial extent of the study area by creating a system of 2 degree x 2 degree tiles which are processed individually through the Daymet software. These tiles are identified by a TileID which is derived within the Daymet algorithm and is consistent throughout the temporal period of the Daymet record. These are the same TileIDs by which the Daily Tile Selection Tool data is made available for download.
The set of ground surface observation stations that are input for the interpolation methods is collected from the heterogeneously spaced stations of the input data from three separate input files of minimum temperature, maximum temperature, and precipitation. The interpolation method at each prediction point is accomplished through an iterative estimation of local station density using the spatial convolution of a truncated Gaussian filter as further described in Thornton et al., 1997. In it, a system is established in which the search radius of stations is reduced in data-rich regions and increased in data-poor regions. This is accomplished by specifying an average number of observations to be included at each point. The search distance of stations is then varied as a smooth function of the local station density. The result is a seamless match of gridded daily data for adjacent tiles.
Daymet Model Outputs and Data Products
For ORNL DAAC distribution, Daymet binary outputs are converted to a spatially referenced, gridded, netCDF file format that follows the netCDF Climate and Forecast (CF) Metadata conventions. The CF standard provides a description of each data variable represented, and the spatial and temporal properties of each 2-degree tile. By providing data in this format, users have a robust way to locate data in both space and time as well as a number of software options for their own research needs. Daily gridded surface data for each of the 7 variables listed below are available:
- maximum temperature
- minimum temperature
- shortwave radiation
- vapor pressure
- snow-water equivalent
The ORNL DAAC also mosaics the 2-degree tiles into a seamless gridded netCDF file of daily data for each variable and each year. These mosaic files, and other pre-derived data products, are available for download through the ORNL DAAC.