Menu Option: Schematic, Catchment Delineation, Load Climate Data, Select Climate Data Source
For catchments in the continental United States, the GridMET climate data source offers two forecasts in addition to its historical data (1979 to yesterday). Both are on the same 4 km grid and carry the same climate variables as the historical GridMET data, so the climate CSV files WEAP creates from them can be joined end to end (see "Chaining historical data and forecasts" below). In the Select Climate Data Source screen, choose GridMET (US only, 1979-yesterday + 28-day and 7-month forecasts, 4 km), then the time period, then the timescale.
The next 28 days, starting tomorrow, at the daily timescale only. It is the average of 48 runs of the NCEP Climate Forecast System version 2 (CFSv2) model, downscaled by the Climatology Lab (University of California, Merced) onto the gridMET grid by applying the model's departures from its 1981-2010 average to the gridMET climatology. A new forecast is produced every day, and WEAP's server processes it every night, so this data source changes daily.
Because it is the average of 48 runs, each day's value is a smooth expectation rather than a weather forecast for that day: the runs agree fairly well on temperature but disagree strongly on the day a storm arrives, so daily precipitation is spread across the period and rarely shows the large one-day totals seen in observed data. The 28-day totals and averages are the meaningful quantities.
The source data provide temperature, precipitation, wind speed, solar radiation, reference evapotranspiration and vapor pressure deficit, but not humidity. WEAP derives the dew point and the minimum, maximum and mean relative humidity from the vapor pressure deficit and the minimum and maximum temperatures, using the same definitions as the historical GridMET data (mean relative humidity is the average of the minimum and the maximum).
The next seven months, starting at the beginning of next month, at the daily and monthly timescales. It is the mean of the North American Multi-Model Ensemble (NMME) seasonal forecast, downscaled by the Climatology Lab onto the gridMET grid. A new forecast is produced each month and processed on WEAP's server in the first week of the month.
The monthly values are the forecast. The daily values are produced by taking the daily pattern of a historical year and rescaling it to match the forecast monthly totals and averages, so the daily sequence (which day it rains, how long a warm spell lasts) is not itself a forecast; use the daily data when your model needs a daily timestep, and read its results at the monthly or seasonal scale.
WEAP downloads the climate data for your catchments once, as one-degree tiles, and keeps them in its Downloads folder (see Folder Locations). For the two forecasts, WEAP checks the server for a newer version each time it needs the data (a few times a day at most), and when the server has a newer forecast it discards the old tiles and downloads the new ones. The climate CSV files in your area are NOT updated automatically: to bring a newer forecast into your model, run Catchment Delineation again for the catchment (or re-run the climate data step for it), which rewrites the CSV files from the current tiles.
The historical GridMET data (1979-yesterday), the 28-day forecast and the 7-month forecast are written to separate CSV files with identical columns, so a single ReadFromFile expression can read them in sequence with the Merge function (the files must have the same columns, in the same order, which these do), for example:
ReadFromFile(Merge(Historical.csv, Forecast28Day.csv, Forecast7Month.csv), 3)
Where the files overlap in time, the earlier file in the list wins, so the observed data are used up to yesterday, the 28-day forecast for the next four weeks, and the 7-month forecast beyond that. The historical file runs to yesterday and the 28-day forecast starts tomorrow, so if today's date falls in the gap WEAP interpolates it from the neighbors, as it does for any missing day.
Both forecasts are produced by the Climatology Lab, University of California at Merced (John Abatzoglou), and distributed through the Northwest Knowledge Network. gridMET: Abatzoglou J. T., Development of gridded surface meteorological data for ecological applications and modelling, International Journal of Climatology (2012), http://dx.doi.org/10.1002/joc.3413. CFSv2: Saha, S., et al. (2014), The NCEP Climate Forecast System Version 2, Journal of Climate, 27, 2185-2208, http://dx.doi.org/10.1175/JCLI-D-12-00823.1. NMME: Kirtman, B. P., et al. (2014), The North American Multimodel Ensemble, Bulletin of the American Meteorological Society, 95, 585-601. See https://climatetoolbox.org/data for the downscaling method.
See also: Catchment Delineation