Merging remote sensing data and national agricultural statistics to model change in irrigated agriculture

Abstract

a b s t r a c tOver 22 million hectares (ha) of U.S. croplands are irrigated. Irrigation is an intensified agricultural land use that increases crop yields and the practice affects water and energy cycles at, above, and below the land surface. Until recently, there has been a scarcity of geospatially detailed information about irrigation that is comprehensive, consistent, and timely to support studies tying agricultural land use change to aquifer water use and other factors. This study shows evidence for a recent overall net expansion of 522 thousand ha across the U.S. (MIrAD-US), was then used to reveal relatively detailed spatial patterns of irrigation change across the nation and the HPA. Causes for the irrigation increase in the HPA are complex, but factors include crop commodity price increases, the corn ethanol industry, and government policies related to water use. Impacts of more irrigation may include shifts in local and regional climate, further groundwater depletion, and increasing crop yields and farm income.Published by Elsevier Ltd.

A multi-scale accuracy assessment of the MODIS irrigated agriculture data-set (MIrAD) for the state of Nebraska, USA

“…The potential exists to develop annual updates of the MIrAD using year-specific MODIS and the USDA crop acreage statistics that reflect the yearly distribution of fields that are irrigated across large areas. Work has recently been completed to develop a 2007 MIrAD applying the same classification strategy to updated MODIS and the USDA Census data for that year and was shown by Brown and Pervez (2014) to be useful for detecting change in the spatial extent of irrigated agriculture. Pervez and Brown (2010) note that 2002 was a relatively dry year across most of the United States with persistent drought conditions over areas such as Nebraska, which provided an enhanced peak NDVI difference between irrigated and rainfed crops that was beneficial for discriminating these two land use classes.…”

Land Use Classification: A Surface Energy Balance and Vegetation Index Application to Map and Monitor Irrigated Lands

“…For that reason, irrigation classification is essentially an arduous procedure since the phenological difference in vegetation canopy is subtle for classification schemes to reliably detect the spectral difference between irrigated and non-irrigated surfaces. Several studies (e.g., [10,12,21]) have determined that the most widely used classification schemes are based on vegetation indices that are only spectrally sensitive to phenological variation under severe conditions such as droughts or desert climate. In this study, the proposed dual index scheme assimilates ETRF (scaled by GI), a phenological and soil water stress index [48] that is sensitive to short and long term water sufficiency or deficiency.…”

“…NDVI has been widely used as an important vegetation and irrigation monitoring tool [8,11,12,[20][21][22]. GI on the other hand, has been a less exploited vegetation index, yet studies [23,24] have found the index more sensitive to chlorophyll than NDVI, Enhanced Vegetation Index (EVI) [25], and Wide Dynamic Range Vegetation Index (WDRVI) [26].…”

Grassland and Cropland Net Ecosystem Production of the U.S. Great Plains: Regression Tree Model Development and Comparative Analysis

“…This effort focused only on grass and non-irrigated cropland, as determined by Brown and Pervez [63], to exclude the substantial production advantages that come with irrigation. To assess normality (a requirement for the application of classical statistics) in regional NEP distributions several approaches were applied: visual inspection of NEP regional frequency plots and the differences between mean and median regional NEP indicated skewness.…”

“…This effort focused only on grass and non-irrigated cropland, as determined by Brown and Pervez [63], to exclude the substantial production advantages that come with irrigation.…”