Climate Driver Analytics for SAHF Member Countries

Precipitation

Per-pixel ENSO precipitation outlook over the SAHF domain (ERA5 reanalysis, 0.25° resolution). Each pixel's color is its historical precipitation response, at this location, to the ENSO phase CPC's official forecast currently favors for this month and the following two. Hover a location for a quick read, or click it for the full basis: how ENSO behaved over the last three months and how rainfall responded here, then what is forecast next, month by month, and the confidence behind it -- plus, further down, that same location's direct ECMWF SEAS5 ensemble forecast for the next 6 months.

Click a location on the map for its outlook.

Precipitation -- Location-specific ENSO Impact outlook

How this map is built, for context alongside the colors above.

Source
ECMWF ERA5 reanalysis, monthly-mean total precipitation (Copernicus CDS API), at native 0.25° resolution -- no spatial averaging into regional boxes, unlike the rest of this pipeline's precip figures.
Spatial span
SAHF domain, 40°N–20°S, 60°E–130°E -- Afghanistan/Pakistan in the west past the Philippines in the east, the Himalayas in the north down through Sri Lanka, the Maldives, and Indonesia past Timor-Leste in the south.
Method
At every pixel, months are grouped by their RONI-classified ENSO phase (El Nino/La Nina/Neutral, the same classification used across this site) and restricted to the 3-month window CPC's forecast currently covers. The map's color and headline badge use that pixel's mean precipitation anomaly (vs. its own calendar-month climatology) across the full 3-month window, tested with a one-sample t-test against zero -- a steadier signal than any single month alone. The click panel additionally breaks this down one calendar month at a time.
Last 3 months
The most recent 3 months ERA5 actually covers. ENSO state comes from RONI; the rainfall figures are observed anomalies at that pixel in each of those exact months, not a model estimate.
Next 3 months
This calendar month plus the following two, per CPC's official probabilistic ENSO forecast (issued monthly). Confidence is High only where CPC assigns at least 70% probability to its favored phase and this pixel's historical composite for that month is statistically significant (p < 0.05); otherwise Low. Single-month confidence is typically lower than the 3-month headline, since a month-by-month breakdown has fewer historical samples to draw on.
Seasonal forecast (ECMWF SEAS5)
A different, direct-ensemble source: ECMWF SEAS5 (Copernicus CDS API, dataset seasonal-monthly-single-levels, originating_centre: ecmwf, system: 51), 51-member ensemble, at 1°×1° resolution, matched to the clicked location's nearest SEAS5 grid cell -- not an ENSO-phase composite like the rest of this panel. The chart shows the same observed months as "Last 3 months" above, then the SEAS5 forecast for the next 6 months as box-and-whisker markers: the 10th–90th percentile (whisker) and 25th–75th percentile (box) spread across the ensemble, with the ensemble mean (not a median) as the center line.

Precipitation -- Region-specific ENSO/IOD/MJO Impact outlook

How the map's colors and detail panel are built, for context alongside the per-pixel map on the other tab.

Source
ECMWF ERA5 reanalysis, monthly-mean total precipitation averaged over each of this site's SAHF precipitation zones, composited against RONI-classified ENSO phase, BOM DMI-classified IOD phase, and ROMI-derived MJO phase group.
Method
At each zone, every plausible combination of the three climate drivers is tested and the best-supported one selected, then checked twice: once by removing the single most extreme month, and again with a randomized reshuffling (permutation) test that screens out results that are simply a byproduct of how long El Nino/IOD episodes tend to last. Zones that don't pass these checks are marked Low confidence; where even an unconfirmed pattern exists, an exploratory lean is shown, clearly labeled as such.
Confidence
High: passed both checks. Medium: directionally consistent but did not clear the permutation-test bar. Low: a weak historical association, or an unvalidated exploratory lean. Full methodology in this report.

2 meter Temperature

Per-pixel ENSO 2m-temperature outlook over the SAHF domain (ERA5 reanalysis, 0.25° resolution), same method as the precipitation map above. Each pixel's color is its historical temperature response, at this location, to the ENSO phase CPC's official forecast currently favors for this month and the following two. Hover a location for a quick read, or click it for the full basis: how ENSO behaved over the last three months and how temperature responded here, then what is forecast next, month by month, and the confidence behind it -- plus, further down, that same location's direct ECMWF SEAS5 ensemble forecast for the next 6 months.

Click a location on the map for its outlook.

2 meter Temperature -- Location-specific ENSO Impact outlook

How this map is built, for context alongside the colors above.

Source
ECMWF ERA5 reanalysis, monthly-mean 2m air temperature (Copernicus CDS API), at native 0.25° resolution -- no spatial averaging into regional boxes.
Spatial span
SAHF domain, 40°N–20°S, 60°E–130°E -- Afghanistan/Pakistan in the west past the Philippines in the east, the Himalayas in the north down through Sri Lanka, the Maldives, and Indonesia past Timor-Leste in the south.
Method
At every pixel, months are grouped by their RONI-classified ENSO phase (El Nino/La Nina/Neutral, the same classification used across this site) and restricted to the 3-month window CPC's forecast currently covers. The map's color and headline badge use that pixel's mean temperature anomaly (vs. its own calendar-month climatology) across the full 3-month window, tested with a one-sample t-test against zero -- a steadier signal than any single month alone. The click panel additionally breaks this down one calendar month at a time.
Last 3 months
The most recent 3 months ERA5 actually covers. ENSO state comes from RONI; the temperature figures are observed anomalies at that pixel in each of those exact months, not a model estimate.
Next 3 months
This calendar month plus the following two, per CPC's official probabilistic ENSO forecast (issued monthly). Confidence is High only where CPC assigns at least 70% probability to its favored phase and this pixel's historical composite for that month is statistically significant (p < 0.05); otherwise Low. Single-month confidence is typically lower than the 3-month headline, since a month-by-month breakdown has fewer historical samples to draw on.
Seasonal forecast (ECMWF SEAS5)
A different, direct-ensemble source: ECMWF SEAS5 (Copernicus CDS API, dataset seasonal-monthly-single-levels, originating_centre: ecmwf, system: 51), 51-member ensemble, at 1°×1° resolution, matched to the clicked location's nearest SEAS5 grid cell -- not an ENSO-phase composite like the rest of this panel. The chart shows the same observed months as "Last 3 months" above, then the SEAS5 forecast for the next 6 months as box-and-whisker markers: the 10th–90th percentile (whisker) and 25th–75th percentile (box) spread across the ensemble, with the ensemble mean (not a median) as the center line.

River discharge

Click a basin on the map. The panel on the right gives its ENSO/IOD/MJO-based outlook: what current and forecast El Niño-Southern Oscillation (ENSO), Indian Ocean Dipole (IOD), and Madden-Julian Oscillation (MJO) conditions mean for discharge here, shown only where the historical record supports a statistically validated relationship. Point color indicates the confidence behind that relationship -- high, medium, or low -- or that none is confirmed; it does not indicate the direction of change. The charts below give the physically-modeled GloFAS discharge forecast for the same basin.

High confidence Medium confidence Low confidence No confirmed relationship
Click a basin on the map.

Colored bands are this basin's own historical discharge climatology (10th-90th percentile), dashed line the median. Solid line is observed discharge; box markers past today are the seasonal forecast, where available.

River discharge -- ENSO/IOD/MJO outlook

How the map's colors and detail panel are built, for context alongside the seasonal/near-term forecast charts below.

Source
This pipeline's own historical GloFAS monthly-mean discharge record at 11 major SAHF river basins, composited against RONI-classified ENSO phase, BOM DMI-classified IOD phase, and ROMI-derived MJO phase group.
Method
At each basin, every plausible combination of the three climate drivers is tested and the best-supported one selected, then checked twice: once by removing the single most extreme month, and again with a randomized reshuffling (permutation) test that screens out results that are simply a byproduct of how long El Nino/IOD episodes tend to last. Basins that don't pass these checks are marked Low confidence; where even an unconfirmed pattern exists, an exploratory lean is shown, clearly labeled as such.
Confidence
High: passed both checks. Medium: directionally consistent but did not clear the permutation-test bar. Low: a weak historical association, or an unvalidated exploratory lean. Full methodology in this report.

River discharge -- Seasonal

How this chart is measured, for context alongside the bands/lines above.

Source
GloFAS historical monthly-mean discharge (this pipeline's own record) for the percentile bands and median; ECMWF SEAS5-forced LISFLOOD 51-member ensemble for the forecast.
Bands
Self-computed monthly percentile bands (10th/25th/75th/90th) from this pipeline's own historical GloFAS monthly-mean record -- an illustrative climatology, not GloFAS's own seasonal-outlook reforecast climatology, which uses a different, proprietary reference period this pipeline doesn't have access to. The dashed line is the monthly climate median.
Observed
Solid line is observed monthly-mean discharge for the last 6 months.
Forecast
Box-and-whisker markers past the dashed "today" line show the GloFAS seasonal ensemble forecast (51 members, ECMWF SEAS5-forced LISFLOOD, issued monthly with about a 4-month lead), where available.

River discharge -- Near-term

How this chart is measured, for context alongside the bands/lines above.

Source
GloFAS daily discharge (this pipeline's own annual-maximum record) for the return-period bands; ECMWF IFS ENS-forced 51-member medium-range ensemble for the forecast.
Bands
Flood return-period thresholds (1.5/2/5/20-year), fit with a Gumbel distribution to this pipeline's own annual-maximum daily GloFAS discharge record (1979-present) -- the same standard method GloFAS's own official threshold grids use, computed independently here, so treat these as a methodologically consistent approximation rather than GloFAS's own published values.
Observed
Black line is observed discharge for the last 10 days.
Forecast
Box-and-whisker markers past the dashed "today" line show the GloFAS medium-range ensemble forecast (51 members: 1 control run + 50 perturbed, ECMWF IFS ENS-forced, about 15 days lead), where available. As of GloFAS v4.4, the official Reporting Points viewer also blends in ECMWF's AIFS Single (AI-based) forecast; that product isn't exposed through this pipeline's bulk data source, so it isn't shown here.

Cyclone activity

Click a sub-basin on the map for its ENSO/IOD/MJO-based genesis outlook: what current and forecast El Niño-Southern Oscillation (ENSO), Indian Ocean Dipole (IOD), and Madden-Julian Oscillation (MJO) conditions mean for cyclone genesis here, shown only where the historical record supports a statistically validated relationship. Color indicates the confidence behind that relationship -- high, medium, or low -- or that none is confirmed; it does not indicate the direction of change.

Four sub-basins are modeled: the Arabian Sea and Bay of Bengal in the North Indian Ocean, and the South China Sea and Philippine Sea in the western North Pacific. The western North Pacific pair matters to the eastern SAHF countries -- South China Sea systems and their remnants drive much of the lower Mekong's wet-season flood peaks -- and the split is itself where the ENSO signal lives there: El Niño shifts mean genesis southeast into the open Philippine Sea, La Niña pulls it northwest toward the Indochina coast.

High confidence Medium confidence Low confidence No confirmed relationship
Click a sub-basin on the map for its outlook.

Storms per season, by basin. Filter by ENSO phase (RONI-based, same classification used across this site) to compare El Nino/La Nina/Neutral seasons -- a storm counts toward whichever phase was active in its genesis month. The two basins are drawn side by side rather than summed: their ENSO responses differ in sign as well as size, so a combined total would cancel the signal.

Cyclone activity -- ENSO/IOD/MJO genesis outlook

How the map's colors and detail panel are built, for context alongside the season-by-season chart below.

Source
NOAA NCEI IBTrACS v04r01 genesis rates at four sub-basins -- Arabian Sea and Bay of Bengal (North Indian Ocean), South China Sea and Philippine Sea (western North Pacific) -- composited against RONI-classified ENSO phase, BOM DMI-classified IOD phase, and ROMI-derived MJO phase group. Each basin is halved at a genesis longitude (77.5°E for the North Indian Ocean, 120°E for the western North Pacific) because the two halves respond to those drivers differently enough that a basin-wide count averages the signal away.
Method
At each sub-basin, every plausible combination of the three climate drivers is tested and the best-supported one selected, then checked twice: once by removing the single most extreme month, and again with a randomized reshuffling (permutation) test that screens out results that are simply a byproduct of how long El Nino/IOD episodes tend to last. Sub-basins that don't pass these checks are marked Low confidence; where even an unconfirmed pattern exists, an exploratory lean is shown, clearly labeled as such.
Confidence
High: passed both checks. Medium: directionally consistent but did not clear the permutation-test bar. Low: a weak historical association, or an unvalidated exploratory lean. Full methodology in this report.
Combined drivers
A sub-basin can select a driver interaction (e.g. IOD and MJO acting together, which the Philippine Sea does) rather than any single driver. That is a real result, but this panel composites one driver at a time and cannot reduce an interaction to a single arrow, so those sub-basins show their long-run genesis rate and say so, instead of a direction.

Cyclone activity -- Season totals

How this chart is measured, for context alongside the bars above.

Source
NOAA NCEI IBTrACS v04r01 best-track tropical cyclone records.
Spatial span
Two basins, drawn as separate bars: the North Indian Ocean (NI -- Arabian Sea + Bay of Bengal) and the western North Pacific (WP -- South China Sea + Philippine Sea). Both records begin in 1980.
Method
One count per storm, bucketed by IBTrACS's own season year and genesis time. Every storm IBTrACS carries for the basin is counted, including tropical depressions that never reach named-storm strength -- which is why the western North Pacific totals run well above the named-typhoon counts usually quoted for that basin.
Phase filter
Uses RONI (Relative Ocean Niño Index), this site's primary ENSO classification everywhere. Each storm is tagged by the RONI phase active in its genesis month; unchecking a phase drops those storms from the bars and recomputes the season totals shown. The basin toggles above it work the same way.