Guide
Guide: how to use ENSOscope
A short walkthrough, from "what is this" to "what do I do with it." No background needed.
ENSOscope in one picture
Using it in five steps
-
See what's coming
Check the ENSO forecast: how strong an El Niño or La Niña is likely, and how many months ahead.
Open the Forecast -
Find your region's risk
Map where the event shifts rainfall, drought and heat, and drill down to your country.
Open Teleconnections -
See what past events did
Replay a past El Niño or La Niña month by month for your country, with the floods, droughts, heat waves and outbreaks recorded at the time, and download the One-page summary or the Full analysis.
Open Event replay -
Weigh the confidence
See how reliable the forecast is at that lead and season, verified against the observed record.
Open Skill -
Act on the lead time
Pre-position supplies, brief teams and trigger anticipatory-action plans in the months of warning. The One-page summary puts the outlook in plain words to share with your team.
The steps, made real: an example, a health team in Kenya
Imagine you are a health team in Kenya and you hear a strong El Niño is forecast. Here is exactly what to do, in about five minutes, and why each step matters.
-
Start with the forecast
Open the Forecast tab. The top line tells you in plain words what is coming, for example a strong El Niño strengthening over the next few months, and the bars show how sure the models are. Why it matters: a big event is likely, so it is worth preparing now.
-
Find Kenya on the map
Open Teleconnections, drag the globe to East Africa, and click Kenya to zoom in. Leave the "Source" on its default, that simply means the map is built from real, observed rainfall records.
-
Check the rain
Keep Wet selected and choose the season that matters to you, SON (the short rains, September to November). Teal means wetter than usual, brown means drier. For you: in a strong El Niño the Kenyan short rains usually turn wetter, which raises the risk of flooding and the disease risk that can follow it.
-
Double-check it against the climate model
Still on Kenya, switch the Source from observations to the climate model, CESM2. The wetter pattern barely changes, which is reassuring, and small robustness dots appear where most events do not agree on the sign of the change, so the signal is not robust on that part of the map. Why: observations give us only a handful of past strong El Niños, too few to be sure; the model turns those into hundreds of simulated events, so it can show where the signal is solid and where it is not. When the observations and the model tell the same story, you can trust it more. (See the two kinds of models below for what CESM2 is and how it differs from the forecast.)
-
Check for dry spells
Switch to Drought / Dry. This shows the longest run of dry days, so you can see whether other seasons bring dry spells that strain water supply and health services.
-
Check the heat
Switch the hazard to Heat stress. This map uses WBGT, which is simply a "feels-like" heat number that combines temperature and humidity. A high value means dangerous heat for the body, which matters most for patients, the elderly and young children.
-
Replay a real El Niño
The maps show what an El Niño usually does. Now see what one actually did: open Event replay for Kenya in the 1997-98 El Niño. Kenya's rainfall climbs far above normal through the short rains of late 1997. The timeline underneath shows a flood on the coast in September 1997 and, under Disease, cholera through 1997 and Rift Valley fever from December 1997, each linked to its source. Move the month slider to follow the season, and compare it with the 2023-24 El Niño side by side. Why it matters: a real event shows when the rain came and what followed, which an average cannot.
Fig. 2. Event replay for Kenya in the 1997-98 El Niño. The Pacific sea surface temperature rising through 1997, and below it Kenya's rainfall with a flood recorded in September 1997. -
Take the downloads
On the same page, download Kenya's One-page summary (PDF or Excel), for hospital and field teams: the coming months in plain words, the forecast of rain, dry days, maximum temperature and heat-stress days region by region, and maps. The Full analysis (PDF or Excel), for analysts, adds the charts, the forecast's skill and what the published research says about El Niño and disease. Both are ready to share with colleagues who will never open the site.
-
Ask: can I trust it?
Open the Skill tab. The grid grades how well the model predicted ENSO in the past: the colours run from pale lavender (no skill) to deep purple, and darker, closer to 1.0, means more reliable. If your lead time sits in the reliable part, you can act with confidence.
-
Act on the lead time
You now know, months ahead, that wetter short rains and flood risk are likely. So you pre-position emergency and clean-water supplies, check drainage and access routes, brief your teams, and trigger your anticipatory-action plan before the season, not during it. That early warning is the whole point of the tool.
Swap "Kenya" and "short rains" for your own context: the same clicks work for any country, season and hazard.
Replaying a past event
The teleconnection maps answer what does an El Niño usually do. The Event replay page answers a different question, and often the more useful one before a season: what did one particular El Niño actually do here, month by month.
Pick an event and a country. You get the sea surface temperature that defined the event, and underneath it rainfall, the heaviest ten-day downpour, the longest dry spell, dry days, maximum temperature, heat stress and warm nights, each as a map and as a line through time.
Three things are drawn on every chart, and they answer three different questions:
- The coloured line is the event you chose.
- The grey lines are every other event of the same kind, lined up on January of their own first year. If the coloured line sits outside the grey band, this event was doing something the others did not. Point at a grey line to see which year it is.
- The dashed grey line is normal: the average of the years with no El Niño and no La Niña. Comparing an El Niño only with other El Niños tells you whether it was worse than the last one. It cannot tell you whether it was worse than normal, which is usually what decides whether to act.
The month slider moves everything at once. The world map, both country panels and the marker on every chart step together, so you can follow a season as it develops rather than reading one average for the whole event. That matters because two seasons often differ less in how much rain fell than in when it fell: rain that arrives early and rain that arrives late produce the same total and very different consequences.
The two country panels sit side by side on one shared colour scale, so you can put your event next to any other and see the difference directly. Point anywhere on a map to read the exact value at that spot. Save as PDF produces a one-page sheet of whatever you are looking at.
Reported disasters and outbreaks. Under the charts, a timeline puts the floods, droughts, heat waves and disease outbreaks recorded in that country on the same months, below the rainfall curve, with a link to the source of each record. Read it with two cautions, which the page also spells out: an empty stretch is often missing reports rather than a quiet year, and a marker inside an El Niño shows that the two happened at the same time, not that one caused the other.
By region. A table under the maps gives the value for each province or region, the number a planner writes down.
Downloads. Next to the country, the One-page summary (for hospital and field teams, PDF and Excel, in English and the country's working language) gives the outlook and the forecast region by region on one page; the Full analysis (for analysts, PDF and Excel) adds maps of the country, charts, and what peer-reviewed research says about El Niño and disease there. Every chart comes with a line on how to read it. Past El Niño seasons are read in three levels: drier or wetter when at least 7 in 10 went that way, tends drier or tends wetter when 6 in 10 did, each with a median more than 5% from normal, and no clear signal otherwise.
Share what you see. The page address keeps the event and the country you chose, so you can send someone straight to the same view.
Rainfall is observed (CHIRPS); heat is from the ERA5 reanalysis. Forecasts are drawn dashed and labelled: where an event is still running, the sea-surface-temperature panel carries the live ENSO forecast, and the rainfall and temperature charts carry the regional forecast (the median of the runs and the middle 80% of them) in the months where a model's past forecasts show skill for that region.
Reading the forecast
The Forecast page opens with a one-line ENSO forecast in plain language, then a table of probability bars, one row per lead time. Lead time is how many months ahead the forecast looks (L1 is the month the forecast is issued, L6 is five months later). Each bar is split by colour into the seven ENSO classes (from extreme La Niña in deep blue to extreme El Niño in deep red); the width of a segment is the share of the model's runs in that class, which is the chance. Two forecast models are shown (ECMWF SEAS5 and Météo-France); when they agree, confidence is higher.
Two kinds of models, two different jobs
It is easy to mix these up, so here is the difference in plain terms.
Forecast models: SEAS5 (ECMWF) and Météo-France System 9. Their job is to predict what ENSO will most likely do over the next one to six months, and how confident to be. That is the Forecast tab: what is coming, and how likely.
Climate model: CESM2, run 50 times over 1850 to 2100. We use 1850 to 2014 as the "historical" period of the model. CESM2 does a different job, and it is not a forecast. It turns the too few observed events into hundreds of simulated ones we can choose from. For each ENSO phase, intensity and season we can see the typical impact pattern across the indices, compare it to the observed events, and check with statistics how robust that pattern really is when the real record has too few events. That is the "check against a model" step on Teleconnections.
Reading the teleconnection maps
On the Teleconnections globe, pick a phase, season and hazard; the map shows how that indicator typically departs from the neutral-season average, the usual level in seasons with neither El Niño nor La Niña. Colour tells you the direction:
"Anomaly" is the difference from the neutral-season average (it isolates the El Niño / La Niña effect); "absolute" is the raw value during that phase. The neutral-season average is the mean of past neutral seasons, one per neutral period, chosen by the same full-season rule as the El Niño and La Niña seasons. The Event replay and the country packs compare with normal instead, the average of every neutral month, so their numbers differ slightly. Robustness dots mark pixels where fewer than 60% of events agree on the direction, so treat dotted areas with caution: the colour there is an average of events that pulled both ways, not a signal you can plan on. The dots appear in the anomaly view only, because they are a statement about a change, and the absolute view shows no change, just the average level during that phase. Click a country to drill all the way down to it: a panel under the map then sets each month of the ENSO forecast beside what past events of the most likely class did to rain in that country, in that month's season. Each map shows one source at a time: observed rainfall (CHIRPS), heat from the ERA5 reanalysis, or the CESM2 climate model run 50 times. Compare them to see whether a pattern holds: the observed record holds only a few strong events, the climate model many.
Reading the skill (how much to trust it)
The Skill page grades how well the model has predicted ENSO in the past. The heatmap shows the correlation for every start month and lead time: 1.0 is perfect, about 0.6 and up is useful, near 0 is no skill. Skill is highest at short lead and fades with lead time and across the boreal-spring barrier. Below it, "the actual track" shows the model line against observations year by year, and the intensity panel asks whether the model gets the strength of an event right, not just its sign.
The words you'll see, in plain English
Every technical term on the platform, explained in one line. You do not need any of these to use the tool.
- El Niño / La Niña (ENSO)
- The natural warming (El Niño) and cooling (La Niña) of the tropical Pacific Ocean. It shifts rainfall and heat worldwide, which is why it is worth watching.
- Teleconnection
- The far-away knock-on effect: how that Pacific change reaches your region as more or less rain, drought or heat.
- Climate model
- A computer simulation of the atmosphere and ocean, a "what if" laboratory used to study how El Niño changes rainfall and heat.
- Ensemble (and "large ensemble")
- Running the model many times from slightly different starts. The spread between runs is the uncertainty; a large ensemble is many runs, which gives lots of El Niño examples to average over.
- CESM2
- The large-ensemble climate model we use. Why: the real world has had only a few strong El Niños, too few to draw a clear map, so the model supplies hundreds of events and the pattern becomes reliable.
- CHIRPS
- A long record of observed rainfall, from satellites and rain gauges. It is what actually fell.
- ERA5
- A reanalysis dataset, not raw observations: a complete, consistent record of past conditions (temperature, humidity, and more), reconstructed by blending a numerical model with worldwide observations. Used here for the heat maps. Daily maximum and minimum temperature cover the whole day, 00 to 24 UTC.
- Why observations and a model together?
- Observations are real but short and noisy for rare events; the model fills the gap with many events. When the two agree, the pattern is trustworthy.
- Rx10day
- The heaviest 10-day rainfall total, a flood-risk indicator.
- CDD (consecutive dry days)
- The longest run of days with no rain, a drought or dry-spell indicator.
- WBGT
- A "feels-like" heat number that combines temperature and humidity. A high value means dangerous heat for the body.
- UTCI
- A similar "feels-like" comfort index.
- Anomaly
- The difference from the usual level without El Niño or La Niña, so it isolates their effect. On the Teleconnections maps the usual level is the neutral-season average (one past neutral season per neutral period); on the Event replay and in the country packs it is normal (the average of every neutral month).
- Lead time
- How many months ahead the forecast looks. L1 is the month the forecast is issued, L6 is five months later.
- Skill / correlation
- How closely past forecasts matched what actually happened, from 0 (no skill) to 1 (perfect).
For the full methods and data sources, see the Methodology page.