Publications and Blogs

Forecasting crops using remote sensing data – An important tool for policy making

By Dripto Mukhopadhyay

When we launched ACRA as a research and analytics firm, one of our major objectives was to help business with data as real time as possible. Apart from whatever available from secondary sources, with quote unquote “error margin”, we worked extensively on remote sensing data and used the same for various estimations especially on spatial consumer demography. We thought of working on agriculture front also using remote sensing data. However, we stopped agriculture related estimations etc. at its nascent stage itself due to multiple reasons, one of which was rare clients in private sector who will show interest in the product. However, it was quite interesting to me to work on a methodology whereby we estimate crop productions etc. using remote sensing data, available secondary and advanced analytics.

Yesterday I reviewed a proposal in the same line for an institute where I am involved as an advisor to research activities. Again the earlier passion got ignited. Though there are different opinions about accuracy of such estimations, I always feel that it can indeed be a significant contribution to research as well as policy making. I just wanted to highlight a few points that seems critical to me from the contribution point of view of estimating crop production with remote sensing data.

  • It is far superior in terms of timeliness since any secondary data on the same comes with a time lag
  • Importantly, with the satellite imageries one can monitor the progress very closely considering natural calamities and also made challenges on various fronts of agriculture production
  • One can forecast crop production for different crops even at land parcel level. It signifies that one can have an estimate for individual farmers also when combined with cadastal or revenue maps
  • This helps the government to remove frauds done by certain section of farmers while they purchase crops grown in other states and sell at a higher MSP
  • It can help managing crop sale at the mandis by farmers since the production records are available to mandi authorities. It helps mandi authorities to allot slots for sale of crops allowing to sell the quantity produced. This is especially true when the allotment is done online, for example as done through Meri Fasal Mera Beyora (MFMB) scheme in Haryana.
  • It also helps the state government to allocate a budget beforehand to be spent on MSP of the crops grown in the state since an advance estimate of production is possible.
  • It also helps in determining the genuine loss by the farmer from insurance point of view also.

There is always a common concern raised, especially on accuracy front of such estimates. However, from methodology perspective if the analytics tools can be used with a cautious care, the glitches will not exceed that we generally see in the secondary data itself. I hope the government will gradually understand the criticality of such estimates in near future and the available remote sensing data can be used for a common good.