2011: Validation of Current Models for Estimating ET
For agricultural use, remote sensing evapotranspiration models are important. However, most remote sensing ET models are not in an operational mode on the internet. The downloading of satellite (SEBAL) and weather data and processing of this data are complicated and time consuming. Consequently, if ET models going to be used by the scientific community it is imperative that the computer processing be automated and user-friendly, i.e., the user just input the dates and locations of interests and the model will output the ET results to the user.
We have developed, and continue to refine, a remote sensing evapotranspiration model (RSET), using satellite data. Challenges inherent in the autmation of SEBAL (including RSET) data include:
- The model requires wet spot (canopy cover) and dry spot (no vegetation) temperatures which represent specific ET conditions to calculate ET.
- Incoming satellite data are for floating locations and the model must determine a fixed location and display it on Google Earth so the end user can associate an ET with a specific area.
Our objective was testing the hypothesis that ET calculated using the RSET automated model has a higher accuracy that manually calculated data. This involves comparing the calcualtions under both stressed and non-stressed tree conditions.
The verification of the model was conducted using data for comparison from an eddy covariance (EC) system to measure fluxes of heat, water vapor, and CO2 into the atmosphere. An Eddy Covariance (EC) tower system was installed in 2005 at a creosote shrub area near Las Cruces, NM. (Fig. 1) The data from that tower is being used to evaluate the remote sensing ET flux under stress ET. A second EC tower system was installed over an almond orchard near Bakersfield, CA (Fig. 2). The site in area in mature almond trees represents a non stressed ET site.The data from this tower is being used to evaluate the remote sensing ET flux under stress ET. Evaluation is ongoing. Currently, the researchers are working with the University of California to evaluate the tool for use by a large almond grower.
The RSET model makes is possible to evaluate remote senses ET calculations over a time series. Modification of the equation and techniques can be conducted quickly and other approaches other than use of SEBAL type calculation can be conducted in the future.
Outreach Webpage for Model
See the webpage that we have developed to provide an indepth presentation & discussion of the model: