A table of sensor specifications can be found in Appendix 3. The primary source of ground truth observations for the CDL products is the USDA Farm Service Agency Command Land Unit program . The FSA CLU comprises digitized polygon boundaries of “semi-permanent ‘fields’” and is a confidential NASS-internal data set. Auxiliary input data sources include the USGS National Elevation Data set , the Multi-Resolution Land Characteristics Consortium National Land Cover Dataset . Prior to 2006, classification was performed using a maximum likelihood classifier in the NASS-internal Peditor program, an image processing software written in Pascal and FORTRAN . Beginning in 2006, Rulequest Research’s See5.0 software was used to create a decision tree classifier. This is applied to the remotely sensed imagery using the MRLC NLCD Mapping Tool and ERDAS Imagine. Accuracy reports are presented in state-level metadata files each annual CDL survey. For supervised classification, ground-truthed observations are defined as polygons, and are subsequently buffered inward by 30 meters. This was done in part to reconcile differences between the different spatial resolutions of the remotely sensed imagery . Prior to 2016, this method of inward-buffering was used for validation and the construction of accuracy reports. However, this excluded edge pixels from the accuracy reports. This resulted in a somewhat inflated accuracy assessments. Starting in 2017, only “unbuffered” accuracy assessments are reported. In 2016, the CDL metadata included both “buffered” and “unbuffered” accuracy reports . Overall accuracy for California FSA crops tend to range between 80 and 90 percent. In order to determine the proportion of daily crop water requirements that were met by direct rainfall,livestock fodder system this study used precipitation observations from the Parameter-elevation Regressions on Independent Slopes Model climate mapping system.
Specifically, 800-m daily precipitation rasters were upscaled to 30 meters, using bilinear interpolation . The PRISM Climate Group at the Northwest Alliance for Computational; Science and Engineering at Oregon State University maintains daily 800-m and 4-km raster datasets of precipitation across the 48 conterminous states, spanning back to 1981. The group also maintains raster datasets of temperature , dewpoint temperature, vapor pressure deficit , and 30-year annual “normals” . PRISM rasters are freely available on the PRISM climate group homepage, with the exception of 800-m monthly and daily data, which must be ordered. PRISM precipitation rasters were commissioned by USDA through the Natural Resources Conservation Service to serve as the official spatial climate data sets of the USDA . PRISM rasters are created at a 30-arcsecond spatial resolution and are also available at a 2.5-arcminute resolution , matching previous USDA-NRCS 1961-1990 climate data sets developed in the 1990s. At its core, PRISM is is an interpolation technique that reproduces the spatial climate patterns of the United States, with a particular emphasis on the effect of elevation and slope on precipitation . The method was originally developed by Dr. Christopher Daly of Oregon State University in an attempt to reproduce the process that climatologists used to construct climate maps of the United States . At its core, the model incorporates data from surface weather stations . PRISM utilizes a linear climate-elevation relationship, rather than a multiple regression model due to difficulties in predicting “complex relationships between multiple independent variables and climate”. Instead, weather station observations are weighted by distance, elevation, coastal proximity, topographic facet, vertical layer, topographic position, and effective terrain . Accuracy estimates using single-deletion jackknife cross validation, leave-one-out cross validation, and a 70% prediction interval have been performed on various revisions to the PRISM method . Regional mean absolute error between predicted and observed precipitation and temperature tend to be similar overall and higher in the physiographically complex western United States.
A 2008 evaluation of PRISM for the central California coast saw good agreement between PRISM, WorldClim, and Daymet temperature observations for the central valley of California .The water footprint incorporates the effect of yields on crop water use. Assuming negligible losses of water, the crop water requirement assumed to be equivalent to the actual crop water use. The resultant water footprints can be considered a “best-case scenario”, as inefficiencies in water distribution and application can only increase the actual crop water use, increasing the blue component of the water footprint. Water footprints are expressed in units of cubic meter of water per metric ton of harvested product. From a resource management perspective, the WF of applied water is most valuable for regions that are predominantly reliant on surface water resources. Total WF figures are presented in Appendix 5, sections G-H. From 2008 to 2015, the blue WF was always orders of magnitude higher than the green WF , further demonstrating the minor role of direct rainfall toward satisfying crop water requirements in California . Across large regional extents, the overall water footprint for most hydrologic regions does not vary much year to year, with the exception of isolated fluctuations driven by changes in reported yield . These fluctuations are also visible in the crop specific annual totals . For example, low mint yields in Shasta County in 2013 inflate the 2013 water footprint for mint . This fluctuation is also visible in the regional WF statistic—due to the low intensity of agriculture in the North Lahontan region , the low yield bias on the WF is visible at the regional scale.A 2018 study by the University of California, Davis compared the consumptive use of water by crops in the Sacramento-San Joaquin Delta of California using seven different crop evpotranspiration models. This “Delta ET” study included methods that were based on crop coefficients and methods which are reliant on remotely sensed satellite measurements.
Monthly crop ET values were published along with the region of interest, for the 2015 and 2016 water years. Overall crop ET observations from this study were compared to the monthly mean of the seven ET models from the Delta ET study. Overall, there was general agreement between this study and the methods detailed in the Delta ET study. This study tended to underestimate crop ET each month by no greater than 50% . The highest proportion of underestimation occurred during the winter months. However, due to the small magnitude of wintertime ET,hydroponic nft gully this only resulted in a 22% cumulative underestimation . The Delta ET models include some land cover classes that this study does not model. The SIMS ET model implemented by the Delta ET study does not model semi agricultural/right-of-way and wet herbaceous/sub-irrigated pasture. The results of this study closely match the monthly results from SIMS within 1%.The water footprint can be thought as a measure of the effectiveness of a unit application of water, given yields as the test for effectiveness. Regionally, it is expected that the highest proportion of crop water use would occur in the intensively-cultivated central valley region . Compared to regions less suitable for agriculture . these regions are exceptional in their overall water use. However, they are not exceptional in the water footprint of agricultural activities . For example, Monterey county contained the largest overall average water footprint of agricultural production, in spite of the possessing a small proportion of overall agricultural water requirement. Among agricultural commodities, average water footprints agree with other assessments in terms of rank order of water footprint and overall crop water requirement. For example, nuts and grasses both have a large water footprint and large crop water requirement, compared to other crops modeled in this study . The large proportional crop water requirement could be function of crop-specific ET characteristics, or it could be an artifact of a large overall cultivated area. However, when compared the proportion of harvested acres, lower WF crops make up a slightly larger portion of cultivated acres than fruits and nuts . Droughts can be used to study the effects of reductions the overall amount of water available in a distribution system. For the drought period starting in 2012, reductions in harvested acres were observed, especially with grasses and some specialty crops.The equivalence of crop water use and crop water requirement was a central assumption in this study. The response of the water footprint under deficit irrigation can be an important topic for future work, as reductions in water use may be less effective from a footprint perspective if yields are dramatically affected. This study explored the distribution of water footprints across the State of California regionally, across different commodities, and across a 7 year period, marked by wet and dry climatic extremes. A model of crop water use was coupled with surveyed observations of precipitation, harvest statistics, and a land cover model. Findings from this study revealed an overall insensitivity of the water footprint to climatic extremes and significant inter-annual variability in the metric .
As a highly derivative metric, the water footprint accumulates errors from all of the data sources used in its calculation. Unreliable yield reports can dramatically change the water footprint, due to the power-law relationship between the water footprint and crop yield. By quantifying the uncertainty of this metric, the water footprint could become even more useful as a decision support tool. However, even exploring the relative proportions of water footprints are useful in defining the conceptual extent of the water-use for a given territory or commodity. Future studies can conduct sensitivity analysis of the metric, to examine which input parameters have the greatest effect on water footprint variability. In the course of this study, a framework was created and implemented in R that allows this analysis to be replicated and run with different inputs. This framework can be utilized in future analyses to compare the footprint metric with the ever improving agricultural methodologies found in California, from modeling irrigation efficiencies, to using improved land use surveys and methods of modeling crop evapotranspiration. The framework can also be applied to different regions, provided that there are harvest and crop ET models which adequately characterize the region. An understanding of the water footprint of agricultural production can provide information to the grower who wishes to maximize the economic return of a given volume of water, the state planner who wishes to maximize utility per unit of water allocated, the national administrator who wishes to understand national risks and strengths, or the informed citizen who wishes to align their consumptive activities with a vision for the conditions conferred to the next generation. This information is a critical component of the continuous motivation to characterize relationships between society and natural resource systems, with the ultimate goal of creating sustainable and resilient social and natural systems.Approximately 5 million workers in the United States work in the agricultural industry with potential exposure to a wide variety of respiratory toxicants. Among the various respiratory hazards not well studied is obstructive and restrictive lung disease caused by inorganic minerals . California agricultural workers have increased respiratory symptoms, decreased respiratory function, and increased mortality rates from chronic pulmonary disease compared with the general population . The agricultural environment of the Central Valley of California places individuals at increased risk of exposure to inorganic particles. This region encompasses a rich farming area as well as extensive urban development. The predominantly dry farming techniques of the Central Valley result in high levels of airborne dust from operations such as field preparation and harvesting of row crops and tree fruits . Although soil consists of a mixture of organic and inorganic materials, potential health effects from the ever-present mineral dusts have been largely overlooked. In theory, soil should contain representative portions of all major mineral classes in the earth’s crust; in reality, most agricultural soils are composed largely of silicate materials and crystalline silica with varying amounts of other mineral classes, depending on the local geologic history. The evidence that mineral dust exposure poses a significant hazard to agricultural workers for interstitial lung disease is based on a handful of case reports, inferences from exposures to mineral dusts in other industries, studies of wild and farm animals exposed to environmental dusts, and toxicologic studies . The deposition and clearance of particles within the respiratory system occurs in an in homogeneous manner.