A. The Land Information System (LIS) is an open source flexible land-surface modeling and data assimilation software framework developed here at NASA Goddard within the Hydrological Sciences Laboratory. GLDAS, NLDAS, NCA-LDAS, and FLDAS are specific uses of the LIS software, and these LDAS projects have produced land-surface forcing data and model output going back many decades and continuing to near real-time. Links to download LDAS datasets are located on the right side of this FAQ. There is no "LIS data" - unless, of course, you use LIS yourself to generate your own!
A. Yes, it is now possible to subset the LDAS datasets by region and/or by variable using the GES DISC's Mirador search tool. For additional help, please see these news articles from the GES DISC about the on-the-fly subset service for the GLDAS and NLDAS datasets.
More recently, the NASA GES DISC has enabled subsetting by region and/or variable by clicking on the "Subset / Get Data" link available under each collection. Click on each LDAS collection under "Get Data" to see this link.
You can also subset by region and/or variable through Giovanni. Users can select a bounding box region or subset using shapefiles, including counties, U.S. states, and watersheds.
A. Two extremely useful tools for processing the LDAS data in GRIB-1 format are WGRIB and grib2ctl.pl. More documentation about GRIB data can be found at NCEP. Climate Data Operators (CDO) are also useful to manipulate and analyze GRIB data. Another useful tool is NCL, which handles GRIB-1/GRIB-2 trivially. NCL has all the GRIB lookup tables built-in. Further, it has two command line utilities: "ncl_filedump", which creates a printed overview of the file contents, and "ncl_convert2nc", which converts GRIB-1/GRIB-2 to netCDF adhering to the Climate and Forecast netCDF convention.
A. The WGRIB tool can easily convert the LDAS GRIB data into ASCII. Here's a sample command, which will dump the first record in the GRIB file named "FILENAME" into an ASCII text file named "output.txt":
wgrib -d 1 FILENAME | wgrib -text -i FILENAME -o output.txt
Please see the WGRIB documentation for additional options and assistance. LDAS datasets can also be obtained in ASCII via the GDS at the GES DISC, as well as by using the time series "data rods" feature.
A. A set of variables from the NLDAS and GLDAS projects are now available as time series. These "data rods" datasets allow the access of the data in a long time series at a single point - without having to download the entire files. Access to the time series data is very quick, and is available as a plot or in ASCII format. A prototype to get the data in netCDF format is also available for testing. Please see the Data Rods News Article for additional information.
A. NASA LDAS datasets are available in the following ways:
- Many GLDAS and NLDAS variables are available over the web as a time series (see previous FAQ).
- All LDAS datasets are available for search, subset, and download through the Mirador service.
- All LDAS datasets are available for online plotting, direct statistical analysis/intercomparison, and download through the Giovanni service. Many variables are currently available, with others being added over time. Users can also select a bounding box region or subset using shapefiles, including counties, U.S. states, and watersheds.
- GLDAS and NLDAS datasets are available in the USGS Geo Data Portal (GDP).
- (see NASA GES DISC story on LDAS datasets available in the USGS GDP)
- GLDAS land water content datasets designed for comparison to GRACE TWS are available.
- NLDAS-2 Forcing, NLDAS-2 Noah, and GLDAS-1 Noah datasets are available in the CUAHSI Hydrologic Information System (HIS).
- GLDAS-1, GLDAS-2 (re-processed), and NLDAS-2 forcing are available in the Google Earth Engine.
- NLDAS-2 precipitation can be used in the EPA BASINS framework.
- (see story on NASA GES DISC EPA BASINS collaboration)
A. Yes, GLDAS-2.0, GLDAS-2.1, FLDAS, and NCA-LDAS datasets are natively in netCDF-4 format. GLDAS-1, NLDAS-1, and NLDAS-2 are natively in GRIB-1 format. However, the NASA GES DISC has enabled on-the-fly conversion from GRIB-1 to netCDF-4 by clicking on the "Subset / Get Data" link available under each collection. Click on each LDAS collection under "Get Data" to see this link.
A. The majority of the fluxes in all of the LDAS datasets are defined as positive in the "traditional" direction. For some more details on this, please see the ALMA (Assistance for Land-surface Modeling Activities) webpages, such as on the sign convention, input, and output.
There are some exceptions to this convention, where the sign of the fluxes in the output were inadvertently reversed. Thus, for these variables within the individual datasets listed below, please reverse the sign to make them conform to the positive "traditional" direction:
In NLDAS-2 Mosaic: EVBSsfc, EVCWsfc, SBSNOsfc [*], SNOHFsfc, TRANSsfc [*] - SBSNOsfc should only be reversed before 9 March 2008. SBSNOsfc data on and after 9 March 2008 does not need to be reversed.
The positive direction of all variables in GLDAS and NLDAS are summarized in this table.
A. Many of the output fields for water amounts (e.g., precipitation, evapotranspiration, soil moisture, etc.) are given in units of [kg m-2]. Many users prefer units of [mm]. If the assumption is made that the density of the water in the soil is 1000 [kg m-3], then the value in [kg m-2] is identical to the value in [mm]:
[kg m-2] ----> [converting to] ----> [mm] kg m^3 1000 mm -- X ----- X ------- = mm m^2 1000 kg 1 m
The density of water does vary slightly with temperature. Users can use the 2-m air temperature or the soil temperatures (as appropriate) to calculate the density of water for each grid box and time period, if desired.
A. Many of the output fields for fluxes (e.g., transpiration, direct evaporation from bare soil, etc.) are given in units of [W m-2]. Many users prefer units of [mm day-1]. If assumptions are made to use the value of the latent heat of vaporization of water at 0 degrees Celsius [2.5x10^6 J kg-1] - and that the water density is 1000 [kg m-3], then the conversion from [W m-2] to [mm day-1] is given as:
[W m-2] ----> [converting to] ----> [mm day-1] W kg J m^3 1000 mm 86400 sec mm --- X ---------- X ----- X ------- X ------- X --------- = --- m^2 2.5x10^6 J W sec 1000 kg 1 m day day
The latent heat of vaporization of water and the density of water both vary slightly with temperature. Users can use the appropriate temperature to calculate these values for each grid box and time period, if desired. Also, note that the latent heat of fusion should be used when converting sublimation.
A. The LDAS datasets currently provide soil moisture values in units of [kg m-2] over the entire thickness of the layer indicated. To convert to units of average volumetric soil moisture [m^3 m-3] over the same thickness, all you need to do is divide the [kg m-2] value by the thickness of the layer in [mm]. The assumption is made that the density of the water in the soil is 1000 [kg m-3].
LDAS SM amount ----> [converting to] ----> LDAS SM volumetric kg m^3 1000 mm 1 -- X ----- X ------- X ----------- = volumetric soil moisture m^2 1000 kg 1 m thickness of layer in mm
As an example, for the 0-10cm layer, just divide by 100mm to convert the data from [kg m-2] to volumetric soil moisture. For the 10-40cm layer, divide by 300mm, and so on for the other soil moisture layers. The volumetric soil moisture values range between wilting point and porosity (e.g., ~0.01 to ~0.5). If we further divide the volumetric soil moisture [m^3 m-3] by porosity [m^3 m-3], we will get the soil moisture fraction which ranges from 0 to 1.
A. All temperature fields in the LDAS datasets are in units of Kelvin. However, some users have reported that when GRIB-1 data is plotted in Python/GDAL, the values appear to be in units of Celsius. The GDAL GRB driver has an option to automatically convert units to "metric" for versions of GDAL >= 1.9.0. In Python, adding the line:
at the beginning of the script will result in temperature units being read in and displayed correctly in Kelvin.