Best Practices for Mission Applications

Emergency driven remote sensing estimate of national crop production losses: 2010 Pakistani flood case study

—Tatiana Nawrocki, USDA

In August of 2010, heavy rains caused widespread flooding along the Indus River in Pakistan. As part of the response ASRC analysts provided imagery analysis support to the USDA Foreign Agriculture Service Office of Global Analysis, which required near real-time reporting of flood impacts on crop production and damage assessments. Multiple sources of remotely sensed data helped to estimate the spatial extent of the flood and to assess damage to agricultural crops in affected regions. Daily MODIS products provided dynamic outlines of the flooded zone along the entire Indus River. The accuracy of these MODIS-derived flood extents was verified using LANDSAT and SPOT 4 imagery in select locations. Kharif season crops (predominantly rice and cotton grown in May - October) were mapped with GIS and a combination of data from NDVI time-series analysis, imagery classification and agricultural statistics. Results of analysis allowed estimating flood-affected agricultural area and potential crop loss. Analysis and evaluation of satellite data resulted in timely updates to crop production forecasts. The accuracy of forecasted crop production figures may be measured when statistical data become available.

This abstract does not represent official policy of the US Department of Agriculture.