The California Irrigation Management Information System (CIMIS) currently manages over 145 active weather stations throughout the state. Archived data is also available for 85 additional stations that have been disconnected from the network for various reasons. Most of the CIMIS stations produce estimates of reference evapotranspiration (ETo) for the station location and their immediate surroundings, often in agricultural areas. Because of California's diverse landmass and climate, many locations within the state lack a representative CIMIS station. Some counties, for example, do not have a CIMIS station and others have only one or two stations. As a result, there are significant spatial ETo data gaps, especially in urban areas. In an attempt to mitigate this problem, CIMIS initiated a study to investigate the possibility of coupling remotely sensed satellite data with point measurements from the CIMIS weather stations to generate spatially distributed ETo values (ETo maps).

A contract was awarded to the University of California Davis (UCD) remote sensing group, lead by Professor Susan Ustin, to conduct the study. The Department of Water Resources (DWR) formed an advisory committee comprised of individuals with expertise in remote sensing, GIS, modeling, and water management from DWR and UCD. The committee met, on an as needed basis to discuss new developments and plan future actions. After thorough research, the team decided to use combinations of data derived from satellites and interpolated from CIMIS station measurements to estimate ETo at a 2 kilometer (km) spatial resolution. The resulting product has been vigorously tested and has demonstrated a degree of accuracy that is acceptable for most irrigation applications. The CIMIS program will continue to evolve and expand to meet the future irrigation information needs of California. For a brief description of the methodology used to generate the ETo maps, see the Spatial Model discussion below.

Daily reference evapotranspiration (ETo) at a 2 km spatial resolution are calculated statewide using the American Society of Civil Engineers version of the Penman-Monteith equation (ASCE-PM). Required input parameters for the ASCE-PM ETo equation are solar radiation, air temperature, relative humidity, and wind speed at two meters height. These parameters are estimated for each 2 km pixel using various methods.

Daily solar radiation is generated from the visible band of the National Oceanic and Atmospheric Administration's (NOAA) Geostationary Operational Environmental Satellite (GOES) using the Heliosat-II model. This model is designed to convert images acquired by the Meteosat satellite into maps of global (direct plus diffused) irradiation received at ground level. The model has also been used with other geostationary satellites such as the GOES. For details on the Heliosat-II model and its accuracy, please refer to the HelioClim web page.

Air temperature, relative humidity, and wind speed values at each pixel were obtained by interpolating point measurements from CIMIS stations. Originally two interpolation methods, Spline and DayMet, were selected based on accuracy of results, code availability, and computational efficiency. Spline – the method currently used – is an interpolation method that fits a surface through or near known points using a function with continuous derivatives. Two- or three-dimensional Spline is used based on which weather parameter is to be interpolated.

The accuracy of both methods has been tested using cross-validation analysis, but DayMet is no longer used. DayMet is an interpolation method that was developed at the University of Montana to generate daily surfaces of temperature, precipitation, humidity, and radiation over large regions of complex terrain. It determines the weights associated with a given weather station for each point where weather parameters are to be determined depending on the distance and density of the stations.

The accuracy of ETo values estimated from these methods depends on many factors. For example, solar radiation remotely sensed through GOES is significantly affected by such factors as cloudiness and snow cover. Therefore, mountainous areas with snow cover and coastal areas with cloud and fog are more susceptible to errors. Also, interpolation accuracy is affected by the density of the weather stations and geographic features of the region. CIMIS stations are purposely placed in irrigated, open, flat areas – usually valleys – to provide the best reference data for adjacent farmlands and other irrigated areas. As a result, interpolation in valleys between CIMIS stations may not provide accurate data for mountainous terrain. Despite these potential problems, however, we believe the ETo estimates provided will be superior to only using data from a distant weather station with a different microclimate. For CIMIS station Siting criteria click the “Siting” tab.

For detailed descriptions of the methodology used to map daily ETo, refer to the RESOURCES navigation button above, click on the Publications tab, select Other Published Articles, and then select the pdf file named "Daily reference evapotranspiration for California using satellite imagery and weather station measurement interpolation" (October 2007 – Civil Engineering and Environmental Systems).

If you would like to get ETo data created using the methods described here, you may do so by logging into your account, clicking on the SPATIAL navigation bar, then on the Spatial Report tab. If you do not have a CIMIS account, you can create one by registering with us. Please note that CIMIS data is provided free of charge and registration is required for statistical purposes only.