SUSTAINABILITY · 9 min read
Sri Lanka Has Fixed Its Maha Planting Date. The Reason It Added Is the One Nobody Was Arguing About
On 4 September, the Department of Agriculture ended a three-way public argument by naming a sowing window: 15 to 25 October, across a target of 830,000 hectares. Through August, the case for moving Maha earlier was to escape a drought. The Department kept that reason—and added a second: get the crop established before a flood. Six days later, the national forecaster published the reading underneath both.
By the ESGNexus Editorial Team · September 2026 · Estimated reading time: 9 minutes
KEY TAKEAWAYS
|
For a fortnight, three public positions emerged on when Sri Lanka should plant its main rice crop, and no one could say which one the government held. The Sunday Times reported on 23 August that farmers were divided over a proposal to bring the 2026/27 Maha season forward by around six weeks to get ahead of a forecast El Niño. The Department of Agriculture had separately indicated an early-October commencement. On 31 August, the Daily FT ran a column by a former Director of the Natural Resources Management Centre arguing the opposite: that cultivation should be delayed until December.
On 4 September, the Department settled it. Farmers are to begin land preparation as soon as possible and complete sowing between 15 and 25 October. The target for the season is about 830,000 hectares of paddy.
That is the news, and it is the smaller half of it. The larger half is the reason the Department gave, because half of it is not the reason anyone was arguing about in August.
What Was Actually Announced
The advisory has four moving parts, and only the first has been widely reported.
| Element | What the Department said |
| Sowing window | Land preparation to begin as soon as possible; sowing to be completed between 15 and 25 October 2026. |
| Why earlier | Sowing early ensures crops are roughly a month old before intense November downpours arrive, preventing flood damage, and allows 3.5-month varieties to be harvested in late January before the dry spell sets in. |
| The 2027 problem | Temperatures are forecast to rise and rainfall to decline in January, February and March 2027. Stricter water discipline is described as vital to a successful Yala season in 2027. |
| Where water is short | Farmers in paddy areas facing water shortages are encouraged to switch to maize, green gram, cowpea or finger millet, and remain eligible for the fertiliser subsidy on paddy land. |
Source: Department of Agriculture advisory, 4 September 2026, as reported by Xinhua and EconomyNext, 4 September 2026. The wording in the second row is EconomyNext’s, quoted in full because the two halves of that sentence are the subject of the next section. No Department press release for the advisory was found on doa.gov.lk or agrimin.gov.lk at the time of writing.
Two officials are named on the record, and both are Directors General, which is what separates this from a seasonal reminder. G G Wickramasinghe, Director General of the Department of Agriculture, set out the irrigation method the Department wants used: alternate wetting and drying. A L K Wijemannage, Director General of the Department of Meteorology, supplied the forecast the advisory sits on.
| Moving into next year — particularly January, February, and March — temperatures will climb further while rainfall drops, ushering in a distinctly dry and warm period. — A L K Wijemannage, Director General, Department of Meteorology, 4 September 2026 |
The Rationale Did Not Invert. It Gained a Half.
Through August, every version of the argument for moving Maha earlier was a drought argument. Plant sooner, the reasoning ran, and the crop is harvested before the El Niño peak dries the tanks. It is a familiar shape in Sri Lanka, and it is the shape the farming press, the agronomic columnists and this publication were all working in.
The Department’s 4 September advisory did not replace that argument. It kept it, and put a second one in front of it. EconomyNext’s report of the briefing carries both in a single sentence: sowing early “ensures crops are roughly a month old before intense November downpours arrive, preventing flood damage and allowing 3.5-month varieties to be harvested in late January before the dry spell sets in”. The reason for the drought is the second half of that sentence. The flood reason is the first—and it is the one nobody in the August argument made.
That distinction decides how the year should be read. The August debate was conducted entirely in the vocabulary of water shortage. The advisory is driven by above-average rainfall in October and November, from El Niño conditions together with a positive Indian Ocean Dipole, and the sowing date it names ensures the crop has roots in the ground before the heavy rain arrives. The word doing the work is not dry. It is November.
Nor has the drought been cancelled — the advisory says where it goes. According to the Department of Meteorology, the dry, warm period arrives in the first quarter of 2027, after this Maha is harvested and immediately before Yala 2027 is planted. A 3.5-month crop sown in the second half of October is off the field before it arrives. That is why the water-conservation half of the advisory reads oddly if you assume it is about this season. It is not. It is about the next one.
For anyone carrying agricultural exposure, this is a materially different year from the one described in August. A drought year concentrates losses in yield and irrigation costs. A wet-then-dry year concentrates it in timing: a crop lost to standing water in November is a total loss on that acreage, and a reservoir drawn down through a dry January is a different problem again, three months later, in a different season and on a different balance sheet.
The National Forecaster Has Now Published Its Reading, and It Says Two Things at Once
The advisory’s stated basis is El Niño conditions plus a positive Indian Ocean Dipole. The Dipole is the second lever, and it is the one worth following, because the Sri Lankan documentary record on it has just moved.
The starting point is domestic and older than the advisory. On 3 July 2026, the Department of Meteorology issued an awareness note, referenced DoM/Awareness/02-2026, which states that “from October onward, rainfall is generally expected to be above normal” and that “if a positive Indian Ocean Dipole (IOD) develops, enhanced rainfall conditions may continue until December”. That is the advisory’s rainfall logic, written down by the national forecaster, two months before the advisory. The same note puts the probability of a very strong El Niño developing between November 2026 and January 2027 at 63 per cent.
The measurement then lagged the scenario. The Seasonal Forecasting Division’s bulletin of 2 August 2026 recorded the weekly Dipole index at minus 0.06 degrees Celsius as of 12 July 2026 — neutral — and a positive phase as a model possibility rather than an observation. On 3 September, the World Meteorological Organisation reported a positive Dipole as expected, at a September-to-November seasonal mean of 0.9 degrees Celsius, and the advisory followed a day later.
On 10 September, the Division published again, under reference SF-2026-09-ENSO-IOD, titled Current condition and forecasts of ENSO and IOD for SON 2026. It is six days younger than the advisory, and it is the document a Sri Lankan planner should now be reading. Two of its sentences matter most, and they pull in different directions on purpose.
| The Indian Ocean Dipole (IOD) is currently neutral. As of 9 August 2026, the weekly IOD index is +0.41 degrees Celsius. This is the third consecutive week that the index has reached the positive IOD threshold (+0.4 degrees Celsius), with sustained values above this threshold required before an event is considered established. — Department of Meteorology, Seasonal Forecasting Division, SF-2026-09-ENSO-IOD, 10 September 2026 |
Read quickly; that looks like a contradiction. It is not, and that distinction is the most useful thing in the document. The index has crossed the positive threshold and has stayed there for three weeks running. The event is not declared, because declaring one requires sustained values over a longer run. The Department is describing a moving measurement and a classification that has not yet changed, which is what a careful forecaster does. Anyone who reports this as “Sri Lanka confirms a positive Indian Ocean Dipole” will have misread it, and anyone who reports it as “the Dipole is still neutral, so the rainfall case is unproven” will have misread it in the other direction.
On El Niño, the same bulletin is less equivocal. It states that “El Niño is strengthening, with a greater than 90% chance of a very strong event during the Northern Hemisphere fall and winter 2026-27”, and that the greater-than-90-per cent likelihood runs across the overlapping seasons from September-November through November-January.
| Question | Dept of Meteorology, Sri Lanka | WMO, published 3 Sep 2026 | The advisory, 4 Sep 2026 |
| El Nino | 3 Jul note: a 63 per cent probability of a very strong event, November 2026 to January 2027. 2 Aug bulletin: an 81 per cent chance of a very strong event, October to December 2026. 10 Sep bulletin: greater than a 90 per cent chance of a very strong event, September-November through November-January 2026-27. | Expected to reach very strong intensity before peaking towards the end of the year.Niñoo 3.4 anomaly around 3.6 degrees Celsius for September to November. Persistence through February 2027 is close to certain. | Prevailing El Niño conditions, cited as a driver of above-average October and November rainfall. |
| Indian Ocean Dipole | 3 Jul note, as a condition: if a positive Dipole develops, enhanced rainfall may continue until December. 2 Aug bulletin, as a measurement: neutral, index at minus 0.06 degrees Celsius as of 12 July. 10 Sep bulletin: still classified neutral, index at plus 0.41 degrees Celsius as of 9 August, a third consecutive week above the positive threshold. | A positive phase is expected to develop, reaching a seasonal mean of 0.9 degrees Celsius from September to November. | A positive Dipole, cited as a co-driver of the above-average rainfall forecast. |
| Currency | The 10 September bulletin post-dates the advisory by six days. The index reading inside it is dated 9 August, roughly a month before the bulletin was issued. | One day old at the date of the advisory. | Reported through two wire services; the Department’s own release was not located. |
Sources: Department of Meteorology, DoM/Awareness/02-2026, Update on El Niño Conditions 02, issued 3 July 2026 — meteo.gov.lk. Department of Meteorology, Seasonal Forecasting Division, Current conditions and forecasts of ENSO and IOD for SON 2026, reference SF-2026-09-ENSO-IOD, issued 10 September 2026 — meteo.gov.lk. The same division’s preceding bulletin, issued 2 August 2026, same location. World Meteorological Organisation, seasonal El Niño update and Global Seasonal Climate Update for September-October-November 2026, both 3 September 2026 — wmo.int.
So the Department of Agriculture was not ahead of its evidence, nor was it departing from its own meteorological service. The rainfall scenario it acted on is the one the Department of Meteorology published in July, and the Division’s own September bulletin has since moved in the same direction on both levers. The worry that has followed this story since August — that a decision covering 830,000 hectares rested on a forecast nobody in Sri Lanka could check — does not survive the documents. The forecast is domestic, dated and public.
One narrower point does survive, and it is worth stating precisely rather than inflating. The index reading published on 10 September is dated 9 August. A bulletin issued in the second week of September, carrying a measurement from the first week of the previous month, is current as a publication and a month behind as an observation. For a farmer choosing between a 15 October and a 25 October sowing date, that gap is immaterial. For an insurer or a lender pricing a November flood exposure, the difference between a live reading and a historical one is the one thing a reader of these documents should know before relying on them.
A Correction We Owe Our Own Readers
A second finding in the Sri Lankan bulletins runs against our previous coverage. This publication has repeatedly cited the Department of Meteorology’s 81 per cent figure as the probability of a strong El Niño, and attributed the possibility of a very strong event separately to the International Research Institute for Climate and Society.
The Department’s bulletin does not say that. It explicitly says there is an 81 per cent chance of a very strong El Niño from October to December 2026, and that the remaining one-in-five chance is that the event will be weaker than very strong. The 81 per cent is a very strong figure. Our earlier framing understated what the national forecaster is actually saying.
The direction of travel is the same in every document we have read. The July note put a very strong event at 63 per cent for November to January; the August bulletin at 81 per cent for October to December; the September bulletin at greater than 90 per cent across September-November through November-January. These are three different windows in three different documents, and they are not a series. A reader who lines them up as 63, then 81, then 90 is reading a trend into figures that answer different questions, and we are not going to print one. What can be said, and all that can be said, is that no document the Department has published points downward. It is worth correcting publicly rather than quietly, because three of our recent articles on reservoirs, generation, and electricity tariffs rest on how severe this season is expected to be.
Why a Wet October Is an Electricity Story as Well as a Farming One
Sri Lanka’s hydrology files are not separable. The same rainfall that decides whether a paddy crop drowns in November decides how much water sits behind the generation reservoirs going into the dry quarter, and that decides how much of the grid runs on imported fuel.
Three live dates make the connection concrete.
- The National System Operator’s assessment, reported on 24 August, was that hydro-generation storage of around 67 per cent was sufficient to support hydropower generation until October. That was a dated claim, and October is the month it was dated to. An above-average October and November change what happens after it expires.
- The Public Utilities Commission’s fourth-quarter 2026 electricity tariff decision falls at the end of September — the first tariff decision taken inside the forecast El Niño peak, and the first real test of whether the regulator prices to the generation reservoirs or to the headline drought.
- The Commission’s August feed-in tariff decision set a March 2027 deadline for adding at least 450 MW of solar photovoltaic with battery storage to the system. That deadline sits precisely in the January-to-March dry window the Department of Meteorology has now described.
This publication has also had to learn one caution twice, and the past fortnight offered a third demonstration. A reservoir percentage means nothing without its population. On 7 September the Irrigation Department reported total reservoir capacity at about 47.3 per cent and, in the same statement, its major reservoirs at about 34.4 per cent, with reservoirs in the Ampara, Batticaloa and Trincomalee districts at around 20 per cent. Those are three different numbers for the same drought, from one department, on one day — and none of them is the hydro-generation figure the power system actually runs on. Any reading of what a wet October does to the grid has to say which reservoirs it refers to.
Source: H.M.P.S.D. Herath, Director of Water Management, Irrigation Department, reported by Xinhua, 7 September 2026. Mr Herath said existing reserves should be sufficient to complete the remaining Yala cultivation period and meet drinking water needs until the next rainfall.
The December Argument Now Has an Answer
The 31 August case for delaying cultivation to December was not careless. A former Director of the Natural Resources Management Centre at the Department of Agriculture argued in the Daily FT that late planting would place the crop’s water demand behind the worst of the dry period.
The Department’s forecast answers it directly, though not by name. A crop sown in December matures across January, February and March — which is exactly the window the Director General of the Department of Meteorology has now described as distinctly dry and warm. On the Department’s own numbers, delaying does not move the crop away from the drought. It moves the crop into it. The October window does the opposite, which is the drought half of the Department’s sentence doing its work.
That does not make the column wrong on its own terms. It was written on 31 August, four days before the advisory, and against the most recent published national measurement at that time, which showed a neutral Dipole at minus 0.06 degrees Celsius. Its author was reading the existing record. The record has since moved twice — once in Geneva on 3 September and once in Colombo on 10 September — and in both instances it has moved against the December case. That is how a well-made argument loses: not because it was careless, but because the measurement caught up with it.
What To Do Now
1. Re-date your climate exposure; do not just re-size it. If your 2026/27 agricultural or hydrological risk register describes a drought in the fourth quarter of 2026, it is now describing the wrong quarter. The Department of Meteorology’s dry period is January to March 2027. Move the date before arguing about the magnitude.
2. Read the 10 September bulletin, and read both of its Dipole sentences. Reference SF-2026-09-ENSO-IOD, published on meteo.gov.lk. It is the current national position. It says the Dipole index has been above the positive threshold for three consecutive weeks, but the Dipole is still classified as neutral. Both halves are load-bearing; quoting either one alone will mislead your board.
3. Ask which reservoirs any percentage refers to. On 7 September, the Irrigation Department reported 47.3 per cent, 34.4 per cent, and around 20 per cent for total storage, major reservoirs, and three eastern districts, respectively. None of those is the hydro-generation figure. A single percentage quoted without its population is not a usable input to a board paper.
4. Treat November as the loss window for standing crops. For plantation, agri-processing, and crop-insurance exposure, the advisory indicates a concentrated flood risk in November for any acreage planted late. Sowing dates in your supply base are now a risk variable, rather than an operational detail.
5. Read the fourth-quarter tariff decision as a climate disclosure. The Public Utilities Commission decides at the end of September, inside the forecast peak. Whatever it decides is evidence about how the regulator reads the season, and it is the cheapest available signal for anyone modelling 2027 energy costs.
What This Article Does Not Claim
It does not claim the Department of Agriculture reversed its reasoning: the drought reason it gave in August is the second half of the sentence it gave in September, and what changed is that it added a flood reason in front of it. It does not claim the Department of Meteorology has contradicted itself: its July note, its August bulletin and its September bulletin answer different questions at different dates, and the September bulletin’s two Dipole sentences are consistent with each other. It does not claim a positive Indian Ocean Dipole has been confirmed or declared — the Department classifies it as neutral, and we report that classification, not around it. It does not claim the Department of Agriculture acted ahead of its evidence or departed from its own meteorological service. It does not claim the forecast is unavailable in Sri Lanka; this publication itself raised that concern in August, and the documents have now answered it. It does not treat 63, 81 and greater than 90 per cent as a trend. It does not claim the 15-to-25 October window is wrong, or that the December argument was. It does not forecast anything itself. And it does not cite any figure. The sources do not state: two wire services reported the advisory from a Department briefing, and while the Department’s own release would be the primary document, we could not locate it.
Sources & Further Reading
Department of Meteorology, Seasonal Forecasting Division, “Current condition and forecasts of ENSO and IOD for SON 2026”, reference SF-2026-09-ENSO-IOD, issued 10 September 2026 — meteo.gov.lk
Department of Meteorology, DoM/Awareness/02-2026, “Update on El Niño Conditions – 02”, issued 3 July 2026 — meteo.gov.lk
Department of Meteorology, Seasonal Forecasting Division, ENSO and IOD bulletin, issued 2 August 2026 — meteo.gov.lk
EconomyNext, “Sri Lanka urges farmers to change cultivation timetable, irrigation due to El Niño”, 4 September 2026 — economynext.com
Xinhua, “Sri Lanka farmers advised to move up cultivation due to El Niño impact”, 4 September 2026 — english. news.cn
World Meteorological Organisation, “El Niño set to become very strong, raising risks of extreme weather into 2027”, 3 September 2026 — wmo.int.
World Meteorological Organisation, Global Seasonal Climate Update for September-October-November 2026, 3 September 2026 — wmo.int
Xinhua, “Sri Lanka reservoir capacity falls to 47.3 pct amid dry weather”, 7 September 2026 — english.news.cn
Daily FT, “Maha 2026/2027: Why delaying cultivation until December may be safer choice”, 31 August 2026 — ft.lk
The Sunday Times reports on farmers’ views on the proposed earlier Maha season, 23 August 2026 — sundaytimes.lk
Asian Mirror, report citing the National System Operator on hydro-generation storage, 24 August 2026 — asianmirror.lk
Public Utilities Commission of Sri Lanka, “Full Final Decision on Feed-in Tariffs, August 2026” — pucsl.gov.lk
| About ESGNexus ESGNexus is Sri Lanka’s independent platform for ESG, CSR, and sustainability intelligence. We track company-level ESG performance, regulatory developments, and sustainability data across Sri Lanka’s listed companies, large unlisted corporates, and state-owned enterprises. All editorial content is independently produced. Sponsored content is clearly labelled. |
Data disclaimer: Information in this article is sourced from publicly available documents. ESGNexus does not independently verify company disclosures. Errors and omissions excepted.