The truth about ground-truthing

Crop water use refers to the amount of water that is lost from vegetation, soil and other surfaces to the atmosphere through evapotranspiration. In this process, the liquid water is released as water vapor. Once it rises into the atmosphere, it becomes invisible, mixes in all directions, thus evading our ability to manage it. Therefore, it is critical that the irrigation is based on evapotranspiration rates while trying to minimize the water loss from liquid to the gas. Thanks to new scientific knowledge, we can now determine evapotranspiration rates in order to develop information for precision irrigation to replace the lost amount of water without the risk of over- or under-irrigation.

The eddy covariance method is based on instrumentation that measures turbulent wind flow (a 3D sonic anemometer) and water vapor concentrations (gas analyzer) directly to determine how the water travels vertically once it turns into vapor. It provides measurements that are continuous, nonintrusive and are representative of the field scale. This stands in contrast to other direct measurement techniques where individual plants and soil samples are used to assume the water status of the whole field, which overlook variations across the field. The accuracy, robustness, precision and reliability of direct measurements make the eddy covariance method essential for evaluating the accuracy and performance of satellite remote sensing, numerical computational models, and other field methodologies that indirectly estimate water loss from crops. Although eddy covariance is the most direct method for obtaining crop water use from the field, it still has around 10% uncertainty, related to instrumental constraints and theoretical assumptions which may not hold in real-world situations. One of the biggest limitations of this method is the cost and complexity of implementation, limiting it to mainly scientific use.

Using eddy covariance

Our University of California-Davis Biomicrometeorology lab makes use of the eddy covariance method, considered by the global scientific community to be the gold standard for direct evapotranspiration measurements, and combines it with modeling and remote sensing. We also develop outreach materials that are oriented in helping growers determine whether or not to irrigate at full irrigation or regulated deficit irrigation.

Much of our effort goes into translating these scientifically obtained datasets into practical suggestions for improving irrigation management. Some of those efforts include determining crop coefficients, calibrating user-friendly models and evaluating remote sensing products. Once we obtain eddy covariance high-frequency data, those are converted to hourly intervals, and our scientific team carefully removes implausible values due to sensor malfunctions and uses the high-quality data to compute daily values of actual ET, or ETa. It is important to mention that reference ET, or ETo, is crucial to upscale the values we measure at a single field to nearby fields with the same crop. The ETo values are freely available in some states that have established a statewide network. In California, that is California Irrigation Management Information System, operated by the California Department of Water Resources.

Once we have ETa measured and quality checked for our local experiments in California, we download ETo from appropriate stations and derive crop coefficients (Kc), which are simply calculated as the ratio of ETa and ETo and published at biomet.ucdavis.edu for local conditions. Crop coefficients give us ETc, which is a crop-specific estimate of ET enabling us to avoid having to directly measure ETa values in each field (ETc is the product of ETo multiplied by Kc). Because the eddy covariance method is pricey, time-consuming, and requires a deep scientific background to maintain it and derive high-quality ET data, CIMIS ETo and Kc values for different crops are useful tools for widespread ET-based irrigation applications. It is also important to note that crop coefficients are not a single value, and they are most useful to be known as a seasonal curve with changing values for the plant development stage, mid-season and leaf senescence. They differ also between the crops as we are trying to use the ETo from short, irrigated grass to estimate crop ETc for different plant architectures (i.e., almond and celery differ quite a bit in their ET and developmental and harvest stages).

Alternative approaches

Sometimes the CIMIS stations are located far away from a particular field, which hinders our capability to derive fully accurate Kc values for each of the crops. Therefore, a lot of effort has gone into estimating ETa using alternative methods.

Satellites are very important remote sensing tool that allow us to estimate ET from large areas. The science behind satellite-based remote sensing is to translate reflected visible light and emitted infrared light from the Earth’s surface into variables that we can then use to estimate ET. Confounding issues arise from the fact that satellite sensors only receive this light once it has passed through the entire atmosphere, which changes it once it has left the surface due to the presence of clouds, haze, varying temperatures and chemical composition. Also, the satellites deployed for this purpose pass overhead once every eight days, which makes hourly to daily ETa estimates using remote sensing less accurate, relying on multiple assumptions that may not hold. Thus, eddy covariance measurements are essential for properly training satellite-based models and informing the numerical modeling community so we can all estimate ET more accurately.

Currently, our lab is actively using EC-based field measurements to verify the accuracy of the data available in satellite-based OpenET products, which is a new web-based platform developed by NASA and USGS for estimating evapotranspiration everywhere at a high spatial resolution. These estimates rely on an approach that accounts for the energy involved in transforming liquid water into vapor released from the field into the atmosphere. OpenET provides evapotranspiration information at the field scale on daily, monthly and yearly intervals, thus promising to help growers better manage irrigation processes, which enables better water accounting and allocation. In the semi-arid Central Valley, we are increasingly experiencing unprecedented levels of heat and drought. Therefore, accurate measurements of evapotranspiration are becoming critical for lowering the expense of producing food and for local agencies to manage scarce water supplies.

Our lab has expertise in all three disciplines of measuring, remote sensing and modeling ET, bridging the gap between shortcomings associated with each method. We are also very enthusiastic about collaborative efforts, where our datasets, collected at around 20 stations across California’s crops, can be put to good use. With these stations’ measurements, with modeling and with remote sensing combined, we can help with other research efforts to determine irrigation needs more accurately. This combined approach will also enable us to scale up ET estimates from a single plot to an entire farm and even to entire regions. Ultimately, this will help growers, scientists and regulators meet food production targets amid declining water resources.

Kosana Suvočarev, PhD, is the assistant professor of cooperative extension, specialist in biometeorology, at University of California, Davis.
Dave Pyles, PhD, is a project scientist in the biomicrometerology lab at University of California, Davis.
Olmo Guerrero Medina is a graduate student researcher in the biomicrometeorology lab at University of California, Davis.
Yifan Guan, PhD, is a postdoctoral fellow between the biomicrometeorology and remote sensing and ecosystem change labs at University of California, Davis.
Jarin Tasnim Anika is a graduate student researcher at the biomicrometeorology lab at University of California, Davis.
Deklan Mengering is a staff research associate in the biomicrometeorology lab at University of California, Davis.
Anushka Kalyan is a high school student in Sacramento, California.