Recent efforts aimed at establishing, standards for data quality indicators and other scoring criteria are driven in part by a desire to properly account for sources of uncertainty in life-cycle assessments . Similar desires have been expressed towards water footprint assessments. As described by Hoekstra : “The field has to mature still in terms of calibrating model results against field data, adding uncertainties to estimates and inter-model comparisons as done in the field of climate studies”. Additionally, researchers now rely on computational methods to synthesize the large quantity of environmental data and observations that are characteristic of studies conducted at large temporal or regional scales. It is still uncommon for data and computational methods to be published along with the completed studies, which obstructs the reproducibility of many hydrologic studies . These later reasons motivated the form of this study—an elementary water footprint analysis decomposed into a reproducible framework. As a case study in resource sustainability, the State of California presents a unique combination of agricultural and economic activities, resource constraints, and environmental monitoring efforts. Among the United States, California has the greatest population, greatest total farm sales , and if considered separately, would rank as the fifth largest economy in the world, by gross domestic product . Nine out of California’s one hundred million acres contain irrigated agriculture; which requires 30 million acre feet of irrigation in an average year, accounting for 80% of the state’s water use . This freshwater requirement is met in part from a vast network of water storage and conveyance infrastructure, which transfer water from the northern third of California,barley fodder system where 2/3 of the precipitation and runoff occurs, to the southern two-thirds, where 3/4 of the anthropogenic water demands are located .
Management of California’s freshwater resources are constrained by dynamic availability on one side and strong, persistent demands on the other. Seasonal variations in precipitation affect the availability of freshwater resources in California The state has recently endured a 5-year drought from 2011-2016, marked by a period from 2012 to 2014 that had the worst drought severity in the past millennium. On the other side, California’s water resources underpin its standing as one of the most productive agricultural exporters in the world and as an important component of the nation’s food security. In 2015, California produced more than 99 percent of the United States’ almonds, pistachios, walnuts, grapes, peaches, and pomegranates . In the same year, international exports accounted for approximately 26 percent of the state’s agricultural production by volume, adding up to 44 percent of the total agricultural sales by value. California is the sole national exporter of many valuable commodities, including almonds, walnuts, and pistachios, which all lie in the top five of the state’s agricultural exports by value . Unpredictable seasonal availability and uncertain international appetite makes it difficult to predict the nature of future constraints and pressures on California’s water resources. There is no guarantee that future climatic, economic, or resource environments will accommodate all of the things that societies value: healthy produce, delicious animal foods, verdant natural vistas, thriving native wildlife, and the autonomy that comes from regional food security. The current attention placed in life-cycle sustainability indicators demonstrates an awareness of the desire to maintain environmental, social, and economic systems without limiting the ability of future generations to meet their needs .
When coupled with scenario analysis, these indicators can support strategic decisions to ensure the security of natural resource supplies. Water footprint assessments have been used to quantify the impact of lifestyles on California’s water resources and have been proposed as policy support tools . Additionally, these assessments have been used to describe the effect of California water resource challenges on international trade networks . While water footprint assessments align with the resource sustainability challenges of California, water scarcity is a problem shared by many nations globally . Therefore, reproducible sustainability assessments are useful in their ability to be applied and compared between different environmental and economic systems.This study used the California Irrigation Management Information System to obtain daily reference evapotranspiration observations across the state. Specifically, the Spatial CIMIS data product was used to obtain raster representations of daily ET0 at a 4 km spatial resolution. This data was upscaled to 30 meters, using bilinear interpolation . The original data is housed and maintained by the California Department of Water Resources , and can be accessed through the CIMIS web interface. CIMIS comprises a network of over 100 automated weather stations that measure the different meteorological parameters at urban and rural sites throughout California. The system was originally established as a project of DWR and the University of California, Davis in 1982 . Each station is sited away from buildings and trees, on a bed of healthy grass that is: “well maintained, properly irrigated and fertilized and mowed or grazed frequently to maintain a height between 10 to 15 centimeters ” . Hourly weather observations are transmitted nightly to Sacramento, where the data are used to compute an average daily evapotranspiration of the reference grass surface underneath each station, using a modified version of the 1977 FAO Penman-Monteith ET0 equation .
The CIMIS Equation differs in its use of a wind function and a method of calculating net radiation from mean hourly solar radiation .The ET0 observations are made publicly available with the primary purpose of aiding agricultural growers develop irrigation schedules. While the CIMIS network provides station-specific ET0 calculations, the Spatial CIMIS data product produces a continuous daily ET0 calculation across the entire state. This is accomplished by using raster observations from the National Oceanic and Atmospheric Administration Geostationary Operational Environmental Satellite system as inputs to the ASCE-Penman-Monteith ET equation . Spatial CIMIS also interpolates temperature and wind measurements from CIMIS stations, to serve as inputs to the ASCE-PM equation . Radiative inputs to the ASCE-PM equation are derived from a clear sky factor that is directly related to cloud cover, as observed by GOES satellite data. Specifically, Spatial CIMIS uses GOES visible imagery to derive a clearness parameter that is directly related to cloud cover in a given grid cell. This is combined with a clear sky solar radiation model developed for the Heliosat-II model . Heliosat-II is a software commissioned by the Solar Radiation Data project, with the purpose of converting images acquired by geostationary meteorological satellites into maps of global solar irradiation, received at ground level . The model incorporates a seasonal turbidity factor, which describes atmospheric attenuation of light due to aerosols and gases. Additional description of inputs to the Spatial CIMIS implementation of the ASCE-PM equation can be found in Appendix. Spatial CIMIS has a weakness in estimating solar radiation in scenarios where changes in the surface albedo can be mistaken for cloud cover. This typically occurs in regions that have snowfall and persistent fog,hydroponic barley fodder system both common winter conditions for some regions in California. Grid cells that contain snow cover and/or fog that persist for greater than 14 days lead to an underestimation of cloud cover and an over-prediction of net radiation during cloudy days Hart et al., 2009. Depending on the location in California, some studies have found good agreement between Spatial CIMIS ET0 and other methods, while others have used Spatial CIMIS after applying correction factors . This study used crop coefficients from Basic Irrigation Scheduling to scale Spatial CIMIS ET0 into crop-specific estimations of evapotranspiration ETc. Kc values for 45 unique crops were selected from the BIS software. These values were supplemented with Kc values from the Consumptive Use Program Plus for garlic and oranges and values from the University of California Division of Agriculture and Natural Resources for some orchard crops. Kc values for peppermint and unspecified caneberries were selected from the AgriMet crop coefficients, which were assembled by the United States Bureau of Reclamation , Pacific Northwest region. Kc values for unstressed Pomegranites were obtained from a study conducted at the Ben-Gurion University of the Negev, Israel. BIS is an application implemented in Microsoft Excel that is used for the planning of irrigation schedules for crops in California .
The software was developed as a collaboration between the University of California, Davis, the California Department of Water Resources, and the University of California Cooperative Extension. The program is currently hosted by the UC Davis Biometerology Group and can be accessed at the BIS home page. Among other uses, BIS is used to determine irrigation schedules, irrigation timings, and maximum allowable soil water depletion for 66 unique crop types. It accomplishes this by estimating crop evapotranspiration given mean climate data for a particular region. BIS partitions evapotranspiration into the component of water evaporated from spoil and plant surfaces and the component transpired by leaves . As the crop matures, the ratio of T to ET increases, until the transpiration component dominates crop ET. To account for the variable ETc , BIS defines: Kc values at different stages in a crop’s life cycle, typical planting and harvest days, and the proportion of the growing period dedicated to each growth stage. These coefficients are defined according to the FAO-56 “single crop coefficient” method, which assigns values according to 4 growth stages of a typical crop: initial growth, crop development, mid-season, and late-season . These growth stages characterize a crop’s daily Kc function, a curve that describes how the values vary as a function of the time in the crop’s growing period. BIS distinguishes between four main crop types.They are characterized by crop coefficients with three inflection points, at 10% ground shading, 75% ground shading, and the onset of senescence. Some type-1 crops such as peas and lettuce, are harvested before their period of senescence. They are characterized by two inflection points, at 10% ground shading and 75% ground shading. Type-2 crops have Kc values that are essentially fixed for most of the season. These include alfalfa, pasture, and most types of turfgrass. Shading of soil by dormant grass may cause an over prediction of soil evaporation and total ETc, however the error may be slight due to the lower overall ETo during the cold winter season Richard L. Snyder, 2014. Type-3 crops do not have a water requirement prior to shoot and leaf growth in the spring and can be characterized by a Kc curve with two inflection points. Type-4 crops represent orchard crops that have fixed Kc values throughout their growing season—similar to type-2 crops. Type-4 crops include subtropical orchards . This study assigned Kc values to individual grid cells according to the crop cover, as observed in the Cropland Data Layer . The United States Department of Agriculture National Agricultural Statistics Service has produced land cover raster image products for major agricultural regions since 1970, and for the 48 conterminous states since 2009 . Annual CDL images can be viewed through CropScape, a web GIS application maintained by USDA-NASS and the Center for Spatial Information Science and Systems at George Mason University. CDL rasters can be downloaded from the CropScape web service, or at the National Resources Conservation Service Geospatial Data Gateway. The CDL was first created by the USDA NASS Research and Development Division, Geospatial Information Branch, Spatial Analysis Research Section . It was based on an image processing and acreage estimation software named Peditor, written in the 1970s and maintained through 2006 . The stated goal of the NASS CDL program is to provide commodity acreage estimates to the Agricultural Statistics Board and other agricultural stakeholders. CDL rasters use standard land cover categories, with an emphasis on agricultural land covers. Records for the State of California begin in the 2007 calendar year; CDL products have a 56-meter spatial resolution from 2007-2009, and a 30-meter spatial resolution from 2009-present. Currently, the CDL is primary constructed from the supervised classification of remotely sensed satellite imagery, from the Advanced Wide Field Sensor onboard the Indian Remote Sensing satellite, RESOURCESAT-1 . This is supplemented with imagery from land imaging sensors onboard the United States Geological Survey Landsat satellites and 16-day Normalized Difference Vegetation Index composites, from the National Aeronautics and Space Administration moderate-resolution imaging spectroradiometer .