Examine how the uneven distribution of monsoon deficits across months and regions undermines the usefulness of an all-India seasonal forecast for agricultural planning.

An all-India seasonal forecast reduces four months of rain over a whole subcontinent to one number. In 2026, an El Niño year, the IMD forecast 90% of the Long Period Average (LPA) [1], and the season ended at 87% [2]. On paper that was close. But farmers plan sowing week by week and district by district, so the number gave them little to act on.

Uneven across months

• June is the kharif sowing window. It received only 65% of LPA, against a forecast of "below 92%" [1][2].
• July recovered to 101% [2]. Late rain cannot fix delayed sowing, so the seasonal total gave false comfort.

Uneven across regions

• East and Northeast India recorded their lowest rainfall since 1901 [2]. Northeast India had been forecast "most likely normal" [1].
• The South Peninsula had its fourth-lowest season since 1901 [2]. This strained the reservoirs that rabi irrigation and drinking water depend on.
• The rain-fed Monsoon Core Zone ended at 94% [2]. One national figure therefore hid both a near-normal core and record-dry outer regions.

Why the all-India number misleads planners

• Wide error band: April's forecast of 92% ±5% [3] covered three categories: deficient, below normal and normal. A result landing inside it proves little.
• Label versus probability: the May update gave a 60% chance of "deficient" but announced the season as "below normal" [1]. Departments act on the label, not the probability table.
• Weak regional skill: the "normal" call for the northeast had only a 35% probability, against 33% for below normal [1]. That is close to a coin toss.
• Scale mismatch: seed reserves, crop switching and drought declarations are decided at district and fortnight scale. A seasonal number does not reach that level.

Way forward

• Make the monthly and regional outlooks and the weekly extended range forecasts [4] the main product, and feed them into district contingency plans.
• When the forecast sits on a category boundary, announce the most likely category in the headline.
• Improve the Monsoon Mission Climate Forecasting System (MMCFS) [1] for sub-seasonal, regional skill. Judge it by whether it changes decisions on the ground.

The national forecast got the direction of the 2026 monsoon right, but the farm losses came from its timing and geography. Advisories by month and district that lead with probabilities would make the IMD's science useful to farmers, advancing SDG 2 (Zero Hunger) and climate-resilient agriculture.