IMD closed the June to September accounting period at 87.4% of the long period average. The south peninsula finished at 76% and east and northeast India at 74%, while the northwest and centre stayed at 94 to 97%. A completed season shows where and when the rain failed, and what a single national number hides.
On 30 September the India Meteorological Department closed the four-month monsoon rainfall accounting period at 87.4% of the long period average, a 12.6% deficit. IMD's report ranks it the fourth lowest since 2001 and the 13th lowest since 1901, and press coverage of the briefing calls it the weakest since 2015. Among IMD's four broad regions, the south peninsula and east and northeast India fell furthest short. One wording point first. 30 September ends the June to September accounting period. It does not mean the monsoon has withdrawn from every part of India, and we come back to that below.
The earlier posts in this series were forecasts and running totals: June, the 15% question in August, and the retreat in September. This one has a completed season, so it can ask three finished questions. Where did the rain fail? When did it fail? And how much of that does El Niño explain? It also scores our own calls against the final number.
IMD's season report, posted on 30 September, gives 759.4 mm of rain against a normal of 868.6 mm, and ranks the season the fourth lowest since 2001 and the 13th lowest since 1901. “Weakest since 2015” is the framing in press coverage of the briefing, including Business Standard and Bloomberg. We have not rebuilt IMD's rankings from its annual series, so they are IMD's, not ours.
Ranking claims across years carry a trap. IMD's own El Niño and monsoon table lists the 2015 June to September departure as -12.7%, a tenth of a point from this year's -12.6%, and Bloomberg cites the same figure. Our August post, quoting an older IMD end-of-season report, gave 2015 as 86% of LPA, a 14% deficit. The historical figures use different reported reference values that have not been reconciled, so we do not compare them directly. On IMD's table, 2026 was roughly as dry as 2015. “Weakest since 2015” is best read as “about level with 2015,” and 2009, at 77% in our August comparison, was drier still, even allowing for the reference-value differences.
The national figure averages two very different seasons. IMD's four-region breakdown, from the regional table in its season report:
| Region | Season rain (mm) | % of LPA | Departure to 23 Sep | Final departure | Change in cumulative departure, points |
|---|---|---|---|---|---|
| Central India | 952.8 | 97% | -5% | -3% | +2 |
| Northwest India | 553 | 94% | -11% | -6% | +5 |
| South Peninsula | 542.0 | 76% | -26% | -24% | +2 |
| East & Northeast India | 1,016.3 | 74% | -26% | -26% | 0 |
| All India | 759.4 | 87.4% | -15% | -12.6% | +2.4 |
The 23 September column is from IMD's extended range outlook issued 24 September. The final columns are from the regional table in IMD's season report of 30 September, which gives season rainfall of 553.0 mm (northwest), 952.8 mm (central), 542.0 mm (south peninsula) and 1,016.3 mm (east and northeast), against long period averages of 587.6, 978.0, 716.2 and 1,367.3 mm. Press coverage agrees, apart from India.com's 2% deficit for central India against IMD's 3%. IMD's report is internally inconsistent in two places: its highlights give 1,017.1 mm for east and northeast India against 1,016.3 mm in the table, and a June narrative figure of 104 mm against 106.8 mm in the monthly table. We use the tables.
Two points stand out. First, the two regions that were furthest behind on 23 September were still furthest behind at the close. The south peninsula's cumulative departure moved two points and east and northeast India's did not move. Northwest India's cumulative departure narrowed most, from -11% to -6%. These are cumulative percentages over different period totals, so they show which regions' season-to-date departures changed, not how much rain each region received in the final week.
Second, this is not a two-corner story. IMD's report rates 17 of 36 meteorological subdivisions, about 42% of the country by area, as deficient and 18 as normal. Odisha, about 5% of the area, was the lone excess. IMD's Director General said at the briefing that Odisha escaped because low-pressure systems kept developing over it. IMD's report counts 14 low-pressure systems, six of them monsoon depressions, and 77 low-pressure-system days against a normal of 57. More systems than normal still produced a dry season, which says where they formed and tracked mattered as much as how many there were.
Month by month, all-India rainfall and the relative Niño 3.4 index moved as follows.
| Month | All-India, % of LPA | Rain vs normal (IMD monthly table) | Niño 3.4 (relative), mean of weekly values | Niño 4 (relative), mean of weekly values |
|---|---|---|---|---|
| June | 65% | 106.8 mm vs 165.3 mm. Provisional on 30 June: 99.5 mm, 39.8% below | +1.0 | +0.7 |
| July | 101% | 283.3 mm vs 280.5 mm | +1.4 | +0.4 |
| August | 84% | 213.3 mm vs 254.9 mm | +1.8 | +0.1 |
| September | 92% | 155.1 mm vs 167.9 mm | +2.1 | 0.0 |
June was the worst month at 65% of LPA, and it came with the weakest Niño 3.4 of the season, about +1.0°C on the relative index. September, the month with the strongest Niño 3.4 at about +2.1°C, recovered to 92%. August, at +1.8°C, was the second-driest at 84%. Four points prove nothing, and monthly rainfall is not expected to track monthly Niño 3.4 one for one. The narrower point is that the national shortfall did not scale with Niño 3.4 strength from month to month. The warming also became more east-heavy: Niño 4 fell from about +0.7°C in June to about zero by September, while Niño 1+2 rose from +2.0°C to +3.9°C. A 2014 NOAA Climate.gov piece by Tom Di Liberto, discussing older events, suggests where the Pacific warms may matter more than how strongly it warms. That is a possibility raised for earlier events, and it does no explanatory work for 2026 here; we show Niño 4 only so the reader can see it.
IMD's report adds two facts about June. Monsoon onset came on 4 June against a forecast of 26 May, plus or minus four days, which IMD says was outside the forecast window and only the second such miss since onset forecasts began in 2005, after 2015. And June recorded no low-pressure-system days, against 24, 27 and 26 in July, August and September. Those are IMD's descriptions of the season. We are not offering them as a cause.
June's number has a footnote. IMD's provisional June figure, reported on 30 June, was 99.5 mm, 39.8% below normal, which is the figure our July post used. IMD's season report gives 106.8 mm, 35.4% below. We have not established the reason for the change, or whether June still ranks among the five driest since 1901.
IMD attributes part of the season to El Niño. Its Director General said the impact was greater in the southern and eastern parts of the country. A senior IMD official told Reuters the shortfall reflected El Niño dampening rainfall. IMD's season report says weak El Niño conditions developed in June and strengthened further, and ETV Bharat reports IMD describing it as strong by the end. NOAA's weekly relative Niño 3.4 rose from +0.7°C to +2.2°C over that stretch.
Attributing the whole shortfall, or its regional pattern, to El Niño runs into three problems.
The fair reading is that El Niño was present, strengthening and named by IMD, and that it does not by itself explain why the south peninsula and the east fell furthest while the central monsoon zone held near normal. As on earlier posts, we are describing shifted odds and not assigning a single cause to a single season. The national number cannot answer the regional question. Only the regional rainfall data, and in time the attribution studies, can.
| Date and post | What we said | Final outcome | Verdict |
|---|---|---|---|
| 7 July, June piece | June 39.8% below normal (IMD provisional, 30 June) | Final June 35.4% below | Matched the source then; IMD's season report gives a different figure |
| 24 August, 15% question | Reuters' unnamed IMD sources said the season might end 15% short; IMD had not published that. An 85% finish would sit one point below 2015 | 87.4%, 12.6% short, inside IMD's own published 29 May forecast of 90% with a 4-point error margin (86 to 94%) | The 15% did not verify. We flagged it as unpublished, and the 85% scenario did not occur |
| 21 September, retreat piece | Confirms if the close is near -15% with the south peninsula 25% or more short. Refutes if the deficit narrowed by more than about four points | -12.6%, south peninsula -24%, narrowing about 2.4 points | Neither condition was met. The two worst regions stayed the two worst; the narrowing was not the size we expected. We do not score the storm's track from final regional totals |
The 21 September call also said the late rain would show up in rabi sowing rather than in kharif output. It is too early to score that part.
A 12.6% national deficit is a true number and an incomplete one. The season's story is that the south and the east fell furthest, June was the worst month, and September recovered. That shape is a more useful test of any explanation than the average, including the one that points at the Pacific.