Embedded Files
Satellite imagery analysis to observe the effect of the drought on agricultural land using ArcGIS
Satellite imagery analysis to observe the effect of the drought on agricultural land using ArcGIS
Project Summary
Between 2011 and 2017 California experienced a period of drought. The project aims to evaluate the effect of the drought on the agricultural crop using NDVI as a proxy for vegetation status. The hypothesis is that after several years of drought, there could be a reduction in cultivated areas.
The data used for the analysis included: satellite scenes of the area acquired by Landsat8 on July7, 2013 and July 13, 2015, the date in which the drought peaked; years 2011 and 2012, were not used in this study because their missing data caused by the Landsat7's SCL failure during those years; California crop mapping from 2014.
Analysis Steps
Composite Band tool to create a single raster from multi-bands including bands 2 (Blue),3 (Green),4 (Red) and 5 (NIR)
Supervised classification to generate 4 classes for each year: water, urban, agricultural land and barren land
The training samples and the Maximum Likelihood Classification method were used to identify the classes on the maps of both years. The histogram was used to assess the separation of the classes
Non-agricultural areas were masked out
NDVI was calculated for both years using the algorithm (NIR – Red)/(NIR + Red)
The classified agricultural area extracted in a previous step was used as mask on the NDVI rasters to isolate the NDVI values for the agricultural areas, then the change NDVI was calculated
The old values, ranging between -0.7 to 0.7, were reclassified as: < 0 into 1 (loss); 0 became 2 (no change); > 0 became 3 (gain)
Results
The analysis shows that the vegetation cover didn't change on the majority of the cultivated land between 2013 and 2015. Actually, a net positive change was observed on approximately 1,000 km2 of the study area.
By not using the scenes closer to the onset of the drought, it is possible that the most drastic effect of the drought was missed. Indeed, it may be safe to assume that the majority of the negative impact could have already happened by the growing season of 2013.
The gain in vegetation observed might be due to mitigation measures employed to counter the drought's negative effects. Indeed, it is plausible that changes to water access, drilling of new wells, introduction of more water-efficient varieties, the response of deciduous plants to stress and their adaptation overtime, may have allowed the vegetation to adjust to the stress.
Further analysis should include a control year as baseline. Also, including data from more years could provide a better picture of the effect of the drought on the agricultural vegetation.
Sources and references
Landsat8 scenes from https://www.usgs.gov/
Land use 2014 – CADWR Land Use Viewer https://gis.water.ca.gov/app/CADWRLandUseViewer/
State Crop Mapping https://data.cnra.ca.gov/dataset/statewide-crop-mapping
California drought comparison by year https://droughtmonitor.unl.edu/Maps/ComparisonSlider.aspx
Page updated
Google Sites
Report abuse