Drought Watch — Kharif 2026
- What this is
- Every mandal through the national drought manual's indicators and triggers, with the declaration calendar.
- What it says now
- 30 mandals read severe and 314 moderate on the manual's indicators; 479 of 615 meet its rainfall trigger.
- What to do with it
- Prepare the drought assessment; it reads the manual, it does not declare.
By the manual’s first trigger, 479 of 615 assessed mandals have had a dry spell. On its three impact indicators, 30 read severe and 314 moderate; 26 more are moderate or severe until a missing indicator decides. This is the evidence the manual asks for, not a declaration: field verification and the notification are the State’s.
Set by 475 dry spells: 4 or more weeks in a row under half the week’s normal rain. On light soils, where the manual allows 3 weeks, 582. By the satellite SPI route, 546.
- Severe30
- Moderate or severe26
- Moderate314
- Normal or moderate19
- Trigger 1 only90
- 1 Aug · early declaration possible
- 31 Oct 2026 · notify kharif drought, with its intensity
- 7 Nov · NDRF memorandum, only if severe
- Up to 3 weeks’ extension for late sowing
Where every mandal stands in the three steps
Step 1 asks whether the rains failed: a dry spell, or a large deficit. Step 2 asks whether it shows, on three of four impact indicators. Step 3, field verification, decides.
Every mandal, and the manual’s matrix filled in for any one of them
Colour the map by the outcome or by any single indicator. Pick a mandal to see each value, the rule it was read against and where it comes from.
A season of dry weeks
The trigger did not arrive at once: it built week by week as the rains failed in one district after another.
How the season dried
Mandals by rainfall so far at the end of each week, and how many had already run four dry weeks in a row (the line). Measured, AP DES gauges.
Show the numbers
| Week ending | Excess | Normal | Deficient | Large deficient | No rain | Dry spell |
|---|---|---|---|---|---|---|
| 11 Oct | 1 | 78 | 450 | 86 | 0 | 475 |
| 4 Oct | 4 | 121 | 413 | 77 | 0 | 475 |
| 27 Sep | 8 | 175 | 374 | 58 | 0 | 474 |
| 20 Sep | 6 | 51 | 452 | 106 | 0 | 474 |
| 13 Sep | 2 | 37 | 401 | 175 | 0 | 472 |
| 6 Sep | 2 | 29 | 372 | 212 | 0 | 466 |
| 30 Aug | 4 | 39 | 382 | 190 | 0 | 420 |
| 23 Aug | 4 | 46 | 384 | 181 | 0 | 382 |
| 16 Aug | 5 | 56 | 385 | 169 | 0 | 373 |
| 9 Aug | 9 | 84 | 374 | 148 | 0 | 370 |
| 2 Aug | 10 | 70 | 356 | 179 | 0 | 365 |
| 26 Jul | 10 | 59 | 308 | 238 | 0 | 283 |
| 19 Jul | 16 | 85 | 325 | 189 | 0 | 189 |
| 12 Jul | 31 | 126 | 284 | 174 | 0 | 129 |
| 5 Jul | 93 | 173 | 223 | 126 | 0 | 37 |
| 28 Jun | 131 | 145 | 216 | 123 | 0 | 0 |
| 21 Jun | 146 | 128 | 187 | 154 | 0 | 0 |
| 14 Jun | 163 | 94 | 164 | 191 | 3 | 0 |
| 7 Jun | 142 | 45 | 128 | 230 | 70 | 0 |
- Ran a dry spell
- 475 of 615 mandals; 582 on the 3-week light-soil rule
- Typical longest spell
- 5 weeks longest: 16 weeks, Naguluppala Padu (Bapatla)
- Driest week
- 97% of mandals under half of normal, 6 Jul–12 Jul
- Short for the season
- 536 mandals deficient or worse to 11 Oct; 86 large deficient
The dry weeks, district by district
Each cell is one week: the darker, the more of the district’s mandals got under half the week’s normal rain. The first week precedes the monsoon’s normal onset and is not counted. Right: mandals with Trigger 1 set.
Which districts carry the most mandals at moderate or worse
Sorted by the share of assessed mandals at moderate or worse; click a heading to sort by an indicator. Medians across each district’s mandals.
| District | Area sown | ||||||
|---|---|---|---|---|---|---|---|
| Bapatla23 mandals · 22 assessed | 20 / 22 | 5 severe · 3 mod–sev · 10 moderate | −46% | 26 | 50% | −0.23 | not shared |
| Annamayya26 mandals · 25 assessed | 25 / 25 | 3 severe · 4 mod–sev · 18 moderate | −58% | 44 | 52% | 0.00 | 26% · reported |
| Kurnool23 mandals · 21 assessed | 19 / 21 | 2 severe · 3 mod–sev · 13 moderate | −51% | 48 | 27% | −0.02 | 90% · reported |
| Sri Sathya Sai32 mandals · 32 assessed | 32 / 32 | 3 severe · 1 mod–sev · 27 moderate | −61% | 39 | 30% | +0.07 | 56% · reported |
| Nandyal29 mandals · 26 assessed | 22 / 26 | 5 severe · 1 mod–sev · 13 moderate | −49% | 41 | 41% | −0.22 | not shared |
| Prakasam27 mandals · 26 assessed | 26 / 26 | 1 severe · 2 mod–sev · 20 moderate | −62% | 37 | 33% | +0.08 | not shared |
| Markapuram21 mandals · 21 assessed | 19 / 21 | 1 mod–sev · 18 moderate | −55% | 32 | 27% | +0.11 | not shared |
| Ananthapuramu33 mandals · 28 assessed | 25 / 28 | 1 severe · 23 moderate | −45% | 40 | 31% | −0.04 | 69% · reported |
| Krishna29 mandals · 22 assessed | 18 / 22 | 2 severe · 1 mod–sev · 12 moderate | −33% | 40 | 62% | −0.20 | not shared |
| Y.S.R Kadapa39 mandals · 36 assessed | 29 / 36 | 5 mod–sev · 21 moderate | −52% | 46 | 33% | +0.01 | 45% · reported |
| Visakhapatnam8 mandals · 7 assessed | 6 / 7 | 1 severe · 3 moderate | −25% | 40 | 65% | −0.29 | not shared |
| Palnadu30 mandals · 23 assessed | 15 / 23 | 2 severe · 13 moderate | −40% | 34 | 32% | −0.11 | not shared |
| Anakapalli26 mandals · 22 assessed | 19 / 22 | 1 severe · 13 moderate | −29% | 50 | 79% | −0.39 | not shared |
| Chittoor24 mandals · 23 assessed | 20 / 23 | 1 severe · 1 mod–sev · 11 moderate | −50% | 53 | 59% | −0.08 | not shared |
| NTR18 mandals · 15 assessed | 13 / 15 | 10 moderate | −38% | 52 | 53% | −0.21 | not shared |
| Guntur15 mandals · 14 assessed | 9 / 14 | 9 moderate | −32% | 34 | 64% | −0.30 | not shared |
| Vizianagaram29 mandals · 26 assessed | 21 / 26 | 1 severe · 11 moderate | −27% | 64 | 57% | −0.07 | not shared |
| East Godavari23 mandals · 21 assessed | 16 / 21 | 1 severe · 8 moderate | −38% | 62 | 57% | −0.17 | not shared |
| Eluru26 mandals · 23 assessed | 18 / 23 | 1 mod–sev · 10 moderate | −31% | 51 | 59% | −0.10 | not shared |
| Sri Potti Sriramulu Nellore32 mandals · 31 assessed | 17 / 31 | 1 mod–sev · 13 moderate | −44% | 50 | 40% | +0.06 | not shared |
| Parvathipuram Manyam13 mandals · 13 assessed | 6 / 13 | 1 severe · 3 moderate | −34% | 59 | 57% | −0.13 | not shared |
| Polavaram12 mandals · 10 assessed | 3 / 10 | 1 mod–sev · 2 moderate | −26% | 49 | 64% | −0.21 | not shared |
| Tirupati33 mandals · 33 assessed | 15 / 33 | 1 mod–sev · 11 moderate | −48% | 53 | 45% | +0.08 | not shared |
| West Godavari19 mandals · 19 assessed | 19 / 19 | 7 moderate | −35% | 62 | 49% | −0.04 | not shared |
| Srikakulam29 mandals · 29 assessed | 23 / 29 | 8 moderate | −19% | 63 | 61% | −0.14 | not shared |
| Kakinada21 mandals · 19 assessed | 11 / 19 | 5 moderate | −25% | 46 | 77% | −0.14 | not shared |
| Konaseema19 mandals · 17 assessed | 11 / 17 | 2 moderate | −31% | 48 | 88% | −0.09 | not shared |
| Alluri Sitharama Raju11 mandals · 11 assessed | 2 / 11 | 0 moderate | −26% | 64 | 72% | −0.15 | not shared |
The two impact indicators that are not per mandal
Reservoir storage and area sown belong to the matrix too, but neither is published per mandal. Both are shown at the level they exist.
Reservoirs against their last ten years
Live storage on 17 Oct 2026 against the average for the same date in 2016–2025. Measured at the dam. The manual applies this only where farming depends on reservoir water; which mandals each one serves is not public.
499 TMC now820 TMC usual39% short · Moderate deficitTable 3.8: storage deficit against the last 10 years' average: under 20% normal, 20–30 mild, 30–40 moderate, 40–60 severe, over 60 extreme. Listed: reservoirs that usually hold 1 TMC or more on this date. The state figure counts all 96 with at least five of the ten years on record.
Fields sown, as reported
Kharif area sown against the season’s normal, 9 Sep 2026: the Commissionerate of Agriculture’s weekly report as quoted by the press (Deccan Chronicle, 11 September 2026). Entered by hand and not refreshed: the department’s site does not answer requests from outside India. Five districts were quoted.
3.2.3.1: sown area under 85% of normal is drought; 75% or less is severe. The manual also asks the State to certify that no further sowing is expected before declaring. Source report.
From evidence to a notification by 31 Oct 2026
The manual gives the State the last word: field checks, the area sown, the irrigated share, and the notification itself.
Steps 1 and 2 for every mandal
Rainfall deviation, dry spells and SPI for Trigger 1; vegetation, soil moisture and groundwater for Trigger 2, with each value’s rule and source. Updated every Monday.
Ground truth in 370 mandals
3.2.6: field checks in 10% of villages, about 5 sites per major crop; crop loss of 33% qualifies, over 50% for severe. Form 11 compiles it village by village.
Area sown, by mandal
The fourth impact indicator, and the certificate that no further sowing is expected. Only district figures for five districts are on this page, as reported.
Irrigated share by mandal
Above 75% irrigated, the State may move a mandal down one rank. The share is not public; canal command areas would also tell which mandals the reservoirs serve.
Notify by 31 Oct 2026
3.4: notify kharif drought by 31 October (extendable by up to 3 weeks for late sowing); NDRF memorandum within a week, if severe.
How this page reads the manual
The manual’s tables, as this page reads themRules, interpretations, sources and what this is not
Read straight from the manual
- Table 3.1 (IMD): normal +19 to −19%, deficient −20 to −59%, large deficient −60 to −99%, no rain −100%.
- 3.2.1 B: consecutive weeks after the monsoon's due onset with under 50% of the week's normal; usually 4 weeks, 3 on light soils.
- Table 3.11: a dry spell sets Trigger 1. Without one, only large-deficient rainfall (or SPI below −1.5) does.
- Table 3.4: VCI 60–100 normal, 40–60 moderate, 0–40 severe. Table 3.6 (2020): PASM 76–100 no drought, 51–75 moderate, 0–50 severe. The State may recalibrate.
- Table 3.9: GWDI above −0.15 normal, to −0.30 mild, to −0.45 moderate, to −0.60 severe, below extreme.
- 3.3.1 Step 2: severe if two of three are severe and the third at least moderate; moderate if two are moderate or worse.
Where the manual had to be read
- 'Moderate' in the severity rule is read as 'moderate or worse'.
- Hydrology bands mild and extreme (Tables 3.8, 3.9) are folded into normal and severe.
- The first week of June precedes the monsoon's normal onset over most of the state and is not counted toward a dry spell.
- GWDI's mean depth is read as the mean for the same month across the record.
Sources, and what each one is
- Rain: AP DES mandal rain gauges via APWRIMS, 1 Jun–11 Oct 2026. Measured.
- SPI: CHIRPS v3.0 monthly, zonal mean per mandal boundary, 2026-06 to 2026-09, against 1981–2025. Satellite estimate.
- VCI: NOAA STAR Blended Vegetation Health Product (VIIRS), VCI, 4 km weekly, weeks of 17 Sep–14 Oct 2026 (dates approximate). 4 km cells over all vegetation, not cropland alone; the manual prefers 56-500 m NDVI/NDWI over agricultural areas. A screening value, not a declaration-grade one.
- PASM: NRSC VIC land-surface model via APWRIMS, plant-available water at 30 cm, weekly values 18 Oct, 11 Oct, 4 Oct, 27 Sep. Modelled.
- GWDI: APWRIMS monthly groundwater depth, the site's own mandal series, 2026-09; same calendar month in every year on record, at least 10 years. Measured, research-access readings.
What this is not
- Not a declaration, and not a recommendation to declare: the manual makes ground truth final and the notification the State’s.
- Prototype mandal boundaries; a few boundaries have no unique gauge or model record and are not assessed.
- The NDVI/NDWI the manual prefers comes from the State Remote Sensing Centre or MNCFC at 56–500 m; the 4 km NOAA index here is a stand-in until that is shared.
- The manual: Manual for Drought Management (updated 2020), Department of Agriculture & Farmers Welfare, Government of India.