Digital Crop Survey: Govt targets to cover all states in Kharif 2025, stresses manual data verification
As the government gears up to cover at least 1 district in non-covered states under the “digital crop survey” (DCS) after 23 crore plots in 3 lakh villages of 435 districts in 14 states are already mapped for the Centre’s ambitious programme, it has been found that there is 19 per cent jump in acreage data from the manual estimate of the crop sown area.
“The overall increase in the surveyed area of 21 districts during last Kharif season was 18.93 per cent higher at 20.09 lakh hectares (lh) in DCS as against 16.8 lh estimated through the conventional system of gathering the crucial information through the village level agricultural and revenue staff,” said Pramod Kumar Meherda, an additional secretary in agriculture ministry.
According to an assessment done by the ministry, there was about 14 per cent jump in paddy acreage from the manually gathered data to the digitally accumulated data in the surveyed districts in Kharif 2024. Similarly, the increase was 48 per cent in sugarcane and 111 per cent in groundnut acreage. Even small millet varieties, normally not given as much importance, like Kakun saw an expansion of 8,500 hectares and Kodo had 32,900 hectares higher area.
“After taking the digital survey, there has to be re-verification of those data manually, particularly wherever discrepancy has been found between the two methods of collecting the information, so that the DCS is made robust,” said an expert adding the manual gathering should also be allowed to continue for next few years until the digital data’s credibility is proven.
“We are actively expanding the DCS, with plans to carry out this ambitious initiative under the Digital Agriculture Mission across all the districts of major states and at least one district in other states during Kharif 2025,” Meherda said adding it will create a comprehensive single source of national crop sown data.
Uttar Pradesh, Bihar, Madhya Pradesh, Tamil Nadu, Andhra Pradesh, Assam and Rajasthan are the best performing states as each of these states have completed over 90 per cent of their targeted plots in the Kharif season last year. Gujarat and Odisha had covered more than 50 per cent of their targeted plots.
He pointed out that traditionally, crop surveys were conducted manually, as part of girdawari involving significant labour, time, and potential for human error as those were typically using paper-based methods that were slow to process and often led to delays in decision-making.
Explaining the future outlook of infusion of technology, he said DCS accumulates a substantial database of crop images, which can serve as training data for machine learning models. “This will facilitate the use of AI to do crop survey, analyse crop health and growth patterns via drones and satellite imagery,” he said adding that these would help improve accuracy as it enables precise measurement of crop acreage.
Union Agriculture Minister Shivraj Singh Chouhan last week told Parliament that mechanisms such as geo-referencing and real-time data information, field-level data capture, supervisor approval and sample check by states help in maintaining accuracy of the data collected under DCS. He said DCS is conducted to collect crop-sown details via a mobile application, ensuring that the data is collected directly from the field.
Stating that elaborate administrative and technical mechanisms have been adopted to ensure the quality and accuracy of data in DCS, he said states are to ensure the authenticity of these data by conducting a minimum of 2 per cent sample check through the State Agricultural Statistics Authority and/or the National Sample Survey Office.
