Revision Date: July, 15, 2016
Please note that this version was superseded by Version 3 on 2016/07/15.
Follow this link to the latest version:
Thornton, P.E., M.M. Thornton, B.W. Mayer, Y. Wei, R. Devarakonda, R.S. Vose, and R.B. Cook. 2016. Daymet: Daily Surface Weather Data on a 1-km Grid for North America, Version 3. ORNL DAAC, Oak Ridge, Tennessee, USA. http://dx.doi.org/10.3334/ORNLDAAC/1328
Contact ORNL DAAC User Services (firstname.lastname@example.org) if you need additional assistance.
This data set provides Daymet output data as mosaicked gridded estimates of daily weather parameters for North America, including continuous surfaces of day length, precipitation, shortwave radiation, snow water equivalent, maximum air temperature, minimum air temperature, and water vapor pressure. The Daymet data product was derived from selected meteorological station data by interpolation and extrapolation algorithms.
The data product covers the period January 1, 1980 to December 31 of the most recently processed calendar year for Canada and the United States, currently 2015. Data for additional calendar years will become available by June of the following year. For Mexico, data are available for the period January 1, 1980 to December 31, 2009.
Data are available on a daily time step at a 1-km x 1-km spatial resolution in Lambert Conformal Conic projection with a spatial extent that covers the conterminous United States, Mexico, and Southern Canada as meteorological station density allows. Data are assembled by parameter and year with each yearly file containing a time dimension of 365 days.
Data are provided in Climate and Forecast (CF) metadata convention compliant (version 1.4) netCDF file formats.
Daymet data are available from the ORNL DAAC via two download mechanisms.
1. Search and Order or Data Browse: Data files may be obtained through DAAC search and order tools or directly by browsing the Daymet data directories. Files are in netCDF version 4.0 format. There are 252 *.nc4 files each containing 7 parameters (dayl, prcp, srad, swe, tmax, tmin, and vp) for 36 years (1980 -2015).
2. THREDDS (Thematic Real-time Environmental Data Services) Data Server: Data can be subset spatially and temporally prior to downloading. THREDDS downloads are available in various formats: netCDF-3 format (*.nc) or netCDF-4 format (*.nc4) if through NetCDF Subset Service and in ASCII or Binary if through OPeNDAP. Subsetting and downloading of files available through THREDDS has a 2-GB file size limitation.
The ORNL DAAC is publishing Version 2.0 of the North American mosaics and will update the Daymet products annually. Version and change history documentation are provided.
ORNL DAAC Version Record:
|Daymet Product Version||ORNL DAAC Release Date||Description|
|Version 3, North American mosaics||July 12, 2016||ORNL DAAC archived and released new daily gridded mosaics for meteorological parameters and updated Daymet Website. All existing Daymet products and tools continue to be available.|
|2015 mosaics||March 7, 2016||ORNL DAAC released daily gridded mosaics for 2015 for each Daymet variable.|
|2014 mosaics||May 18, 2015||ORNL DAAC released daily gridded mosaics for 2014 for each Daymet variable.|
|2013 mosaics||June 13, 2014||ORNL DAAC released daily gridded mosaics for 2013 for each Daymet variable.|
|Version 2, North American mosaics||May 15, 2014||ORNL DAAC archived and released new daily gridded mosaics for meteorological parameters and updated Daymet Website. All existing Daymet products and tools continue to be available.|
|May 10, 2012||NACP Modeling and Synthesis Thematic Data Center (MAST-DC) at ORNL released new daily gridded meteorological parameter data for 2-degree Daymet tiles through the Daymet Website.|
The THREDDS Data Server allows users to find and access data sets of interest from within a simple, hierarchical catalog within a Web browser or compatible client software. Data can be subset spatially and temporally prior to downloading. THREDDS supports data downloads in various data formats. Subsetting and downloading of files available through THREDDS has a 2-GB file size limitation.
Companion Documentation for this Data Set:
Daymet_mosaics.pdf (this user’s guide)
Additional Data Services at the Daymet Project Web Site:
In addition to the Search and Order/Data Browse and THREDDS data download capabilities, two services are available to provide Daymet data in other forms.
Single 1-km by 1-km Pixel Data Extraction Tool
This tool allows a user to obtain daily data for all or a selection of the parameters for selected year(s) for a single geographic point (1-km x 1-km pixel) in a tabular format output file. Direct link: http://daymet.ornl.gov/singlepixel.html
Daymet Tile (2-degree x 2-degree ) of Daily Gridded Surface Data Selection Tool
This tool allows a user to obtain gridded data for the 2-degree by 2-degree tiles that are produced during the Daymet processing routine. Users can download the daily gridded surface data for all Daymet parameters for the selected tile. Daily data for each parameter are distributed as individual files in a netCDF-3 format (*.nc) file format. Direct link: http://daymet.ornl.gov/gridded.html
Cite this data set as follows:
Thornton, P.E., M.M. Thornton, B.W. Mayer, N. Wilhelmi, Y. Wei, R. Devarakonda, and R.B. Cook. 2014. Daymet: Daily Surface Weather Data on a 1-km Grid for North America, Version 2. ORNL DAAC, Oak Ridge, Tennessee, USA. Accessed Month DD, YYYY. Time period: YYYY-MM-DD to YYYY-MM-DD. Spatial range: N=DD.DD, S=DD.DD, E=DDD.DD, W=DDD.DD. http://dx.doi.org/10.3334/ORNLDAAC/1219
For citing specific Daymet downloaded data and subsets used in your analyses and publications document as appropriate:
Date accessed: Date that you downloaded data to account for possible updates within Version 2.
Temporal range: Range of data in year/month/day (YYYY/MM/DD).
Spatial range: Define the bounding box for the data as the Northern- and Southern-most latitudes and Eastern- and Western-most longitudes in decimal degrees. South latitude and West longitude are (-) negative values. Add decimal places as needed for precision.
In addition to the data set citation, the following should be used as the general reference for the methods used to generate Daymet data products:
Thornton, P.E., Running, S.W., White, M.A. 1997. Generating surfaces of daily meteorological variables over large regions of complex terrain. Journal of Hydrology 190: 214 - 251. http://dx.doi.org/10.1016/S0022-1694(96)03128-9
For applications of the radiation and humidity data please include the following citations in addition to the general citation:
Thornton, P.E., H. Hasenauer, and M.A. White. 2000. Simultaneous estimation of daily solar radiation and humidity from observed temperature and precipitation: An application over complex terrain in Austria. Agricultural and Forest Meteorology 104:255-271. http://dx.doi.org/10.1016/S0168-1923(00)00170-2
Thornton, P.E. and S.W. Running. 1999. An improved algorithm for estimating incident daily solar radiation from measurements of temperature, humidity, and precipitation. Agriculture and Forest Meteorology. 93:211-228. http://dx.doi.org/10.1016/S0168-1923(98)00126-9
Daymet is model-produced gridded estimates of daily weather parameters based on daily meteorological observations. The algorithms and software that generate Daymet data products were developed to fulfill the need for continuous surfaces of daily weather data necessary for plant growth model inputs. These Daymet data also have broad applications over a wide variety of scientific and research fields including hydrology, terrestrial vegetation growth models, carbon cycle science, and regional to large scale climate change analysis. 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 in Lambert Conformal Conic projection over the conterminous United States, Mexico, and Southern Canada. The required model inputs include a digital elevation model and observations of maximum temperature, minimum temperature, and precipitation from ground-based meteorological stations.
Mosaicked daily gridded data are available for the seven output parameters and distributed as individual files per year each in a CF compliant (version 1.4) netCDF file format. The netCDF file format is self-describing and compliant to the CF metadata conventions. Each parameter, as well as the spatial and temporal properties of the data, is defined within the header file.
Data are available for each of the seven parameters on a daily time step at a 1-km x 1-km spatial resolution for the conterminous United States, Mexico, and Southern Canada. The spatial extent is the same for all of the files. See the data spatial properties description below.
The data set begins in January 1, 1980 and ends in December of the most recently processed calendar year -- currently 2015. For Mexico, data are available for the period January 1, 1980 to December 31, 2009.
DAAC HTTP Site:
Data are assembled by parameter and year with each yearly *.nc4 file containing a time dimension of 365 days. Files are in CF-compliant netCDF-4 format.
Files are assembled by parameter and year in a flat structure. Filenames follow this syntax:
Where pppp is the respective parameter abbreviation (dayl, prcp, srad, swe, tmax, tmin, and vp) and YYYY is year. For example, precipitation data are provided in 36 files starting with prcp_1980.nc4 through prcp_2015.nc4.
THREDDS Data Server:
Data are presented by year (1980-2015) and then parameter.
Data can be subset spatially and temporally prior to downloading. THREDDS downloads are available in various formats: netCDF-3 format (*.nc) or netCDF-4 format (*.nc4) if through NetCDF Subset Service and in ASCII or Binary if through OPeNDAP. Subsetting and downloading of files available through THREDDS has a 2-GB file size limitation.
Data User Note: The data files on the THREDDS Data Server are the same as on the HTTP site and are individually named the same (e.g., prcp_2012.nc4). The current THREEDS Data Server NetCDF Subset Service provides output files in both netCDF-3 (*.nc) and netCDF-4 (*.nc4) formats.
Parameters, Parameter abbreviations, Units, and Descriptions:
|Day length||dayl||s/day||Duration of the daylight period in seconds per day. This calculation is based on the period of the day during which the sun is above a hypothetical flat horizon|
|Precipitation||prcp||mm/day||Daily total precipitation in millimeters per day, sum of all forms converted to water-equivalent. Precipitation occurrence on any given day may be ascertained.|
|Shortwave radiation||srad||W/m2||Incident shortwave radiation flux density in watts per square meter, taken as an average over the daylight period of the day. NOTE: Daily total radiation (MJ/m2/day) can be calculated as follows: ((srad (W/m2) * dayl (s/day)) / l,000,000)|
|Snow water equivalent||swe||kg/m2||Snow water equivalent in kilograms per square meter. The amount of water contained within the snowpack.|
|Maximum air temperature||tmax||degrees C||Daily maximum 2-meter air temperature in degrees Celsius.|
|Minimum air temperature||tmin||degrees C||Daily minimum 2-meter air temperature in degrees Celsius.|
|Water vapor pressure||vp||Pa||Water vapor pressure in pascals. Daily average partial pressure of water vapor.|
The data are stored and distributed as individual CF-Compliant netCDF file for each parameter. The most current Daymet data are being delivered to the user in terms of both Daymet software and Daymet data versions. Version information is recorded in the header file of each of the CF-netCDF files within the Global Attribute fields; Version_software and Version_data. All Daymet data are provisional and subject to revision.
The Daymet Calendar:
The Daymet calendar is based on a standard calendar year. All Daymet years, including leap years, have 1 - 365 days. For leap years, the Daymet database includes leap day. Values for December 31 are discarded from leap years to maintain a 365-day year.
Spatial Data Properties:
Spatial Representation Type: Raster
Pixel Depth: 32 bit
Pixel Type: float
Number of Bands: 365
Band Information: time
Raster Format: netCDF
Source Type: continuous
No Data Value: -9999
Scale Factor: none
Endian Type: NA
Number Columns: 5,268
Column Resolution: 1,000 meter
Number Rows: 4,823
Row Resolution: 1,000 meter
Extent in the items coordinate system
xll corner: -2015000
yll corner: -3037000
Cell Geometry: area
Point in Pixel: corner
Spatial Reference Properties:
Geographic Coordinate Reference: WGS_1984
Projection: Lambert Conformal Conic
The North American Daymet projection system and parameters:
Projection System: Lambert Conformal Conic
projection units: meters
datum (spheroid): WGS_84
1st standard parallel: 25 deg N
2nd standard parallel: 60 deg N
Central meridian: -100 deg (W)
Latitude of origin: 42.5 deg N
false easting: 0
false northing: 0
Site boundaries:(All latitude and longitude given in degrees and fractions)
|Site (Region)||Westernmost Longitude||Easternmost Longitude||Northernmost Latitude||Southernmost Latitude||Geodetic Datum|
• The data set covers the period January 1, 1980 to December 31 of the most recent processed calendar year. Data for additional calendar years will become available by June of the following year. For Mexico, data are available only for the period January 1, 1980 to December 31, 2009.
The Daymet data have broad applications over a wide variety of research fields including hydrology, terrestrial vegetation growth models, carbon cycle science, and regional to large scale climate change analysis. 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. The continuous gridded surfaces of the Daymet data set were developed to overcome these limitations.
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. 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 average number of stations (n) for temperature extrema is 25; for precipitation, n = 15. 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.
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. The Daymet data set includes estimated SWE as an output parameter 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.
In the Daymet data processing, binary output files from the Daymet model are passed through a number of QA/QC checks that include annual and monthly climate summary evaluations and minimum/maximum value checks. 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” that 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. There are a small number of Daymet tiles that do not run properly through the Daymet algorithm or that do run but have been removed from the Daymet data collection due to poor quality results for all or some of the associated Daymet parameters. The cause of the failure of these tiles is almost exclusively due to low surface observation station density in areas that are sparsely populated.
No Data Tiles
In order to maintain the integrity of the Daymet product, any tile that was not produced because of quality reasons (i.e., low station density) has been replaced with NoData netCDF files for each of the parameters associated with that tile. For these tiles that were not produced, there are NoData netCDF files for each of the parameters: dayl.nc, prcp.nc, srad.nc, swe.nc, tmax.nc, tmin.nc, and vp.nc. These NoData netCDF files will have the appropriate spatial and temporal extents for its specific tile location, but contain only the NoData Value of -9999.
A list of Daymet NoData Tiles by year can be found at: http://daymet.ornl.gov/datasupport.html.
Note that meteorological station data for Mexico is available only for the period January 1980 through December 2009.
Snow Water Equivalent Clarification
Snowpack, quantified as snow water equivalent (SWE), 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 (v2.0) data set includes estimated SWE as an output parameter 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. An important caveat in the use of SWE from the Daymet (v2.0) data set is that the algorithm used to estimate SWE is executed with only a single calendar year of primary surface weather inputs (daily maximum and minimum temperature and daily total precipitation) available for the estimation of a corresponding calendar year of snowpack. Since northern hemisphere snowpack accumulation is commonly underway already at the beginning of the calendar year, the SWE algorithm uses data from a single calendar year to make a two-year sequence of temperature and precipitation, then predicts the evolution of snowpack over this two-year period to provide an estimate of year day 365 (December 31 for non-leap years) snowpack as an initial condition for the January 1 time step of the actual calendar year. The problem with this approach is that it ignores the dependence of January 1 snowpack on preceding calendar year temperature and precipitation conditions, and so generates potential biases in mid-season snowpack which can propagate to biases in late-season timing of snow melt. A better approach would be to calculate snowpack from the continuous stream of calendar years. We intend to apply this improvement in subsequent Daymet releases.
The Daymet model requires spatially referenced ground observations of daily maximum and minimum temperature and precipitation. These observations have been obtained from a number of sources throughout this current Daymet campaign. The ground observations for the United States came from two main sources. The first is the Cooperative Summary of the Day network of weather stations archived and distributed by the National Climate Data Center (NCDC). These data have recently come under the umbrella and are distributed as part of the Global Historical Climatology Network (GHCN)-Daily data set. The second source of surface observation data for the United States is the SNOwpack and TELemetry (SNOTEL) data set managed and distributed by the Natural Resources Conservation Service (NRCS). These stations are primarily in high elevation regions in the Western US and Alaska and are principle in maintaining critical snow pack information. Canadian surface observations were provided by the Government of Canada (Environment Canada) and through the GHCN-Daily data set. The Servicio Meteorológico Nacional provided surface weather observations within Mexico.
Additional inputs for the Daymet algorithm are a digital elevation model (DEM) and Land Mask. The DEM used in this version of Daymet is a North American subset of the NASA SRTM near-global 30 arc second DEM. This DEM was reprojected and resampled from a geographic coordinate system (GCS_WGS_84) to the Daymet Lambert Conformal Conic projection as outlined below. The resampling method used a cubic convolution interpolation with an output cell size set to 1,000 m and a file extent as recorded below. Though no Daymet data is currently available for regions north of 60 degrees North, the SRTM DEM was augmented with GTOPO 30 data above this latitude. Slope, aspect, and horizon grids are derived from the DEM within the Daymet algorithm. A Daymet North American land “mask” file was derived in order to allow Daymet processing to occur for “land” pixels including shallow inland water ways, coastlines, and lake shorelines while excluding only very large bodies of water such as the Great Lakes. The SRTM DEM included the Great Lakes and tended to over-estimate large river and coastal water and was therefore not a good candidate for deriving the Daymet land mask. A reclassification was performed on the MODIS Nadir BRDF-Adjusted Reflectances (NBAR) MODIS land-water mask (Salomon, 2004) as outlined below.
|Old Classification||New Classification|
|0 – Shallow Ocean||Water|
|1 - Land||Land|
|2 – Shallow Inland Water||Land|
|3 – Ocean Coastline and Lake Shorelines||Land|
|5 – Deep Inland Water||Land|
|6 – Shallow Ocean||Water|
|7 – Deep Ocean||Water|
This data is available through the Oak Ridge National Laboratory (ORNL) Distributed Active Archive Center (DAAC).
Telephone: +1 (865) 241-3952
Thornton, P.E., S.W. Running, and M.A. White. 1997. Generating surfaces of daily meteorological variables over large regions of complex terrain. Journal of Hydrology 190: 214 - 251. http://dx.doi.org/10.1016/S0022-1694(96)03128-9
Thornton, P.E., H. Hasenauer, and M.A. White. 2000. Simultaneous estimation of daily solar radiation and humidity from observed temperature and precipitation: An application over complex terrain in Austria. Agriculturaland Forest Meteorology 104:255 - 271. http://dx.doi.org/10.1016/S0168-1923(00)00170-2
Thornton, P.E. and S.W. Running. 1999. An improved algorithm for estimating incident daily solar radiation from measurements of temperature, humidity, and precipitation. Agriculture and Forest Meteorology. 93:211 - 228. http://dx.doi.org/10.1016/S0168-1923(98)00126-9