flood · RwandaWestern Province · signal: tail

Rwanda floods / landslides — May 2023

  1. Act I — Understanding the eventWhat is happening?
  2. Act II — Belief updating & riskWhat do we think is happening?
  3. Act III — Decision & reflectionWhat should we do and why?

Round 1 / 3: Early monitoring (2023-04-25) (2023-04-25)

Event window: 2023-04-252023-05-05 (daily)

Engine state: Antecedent moisture building; ensemble exceedance still modest

Western Rwanda, long-rains. Steep terrain prone to flash floods and landslides; hillside communities exposed. You are the DOC duty analyst as the forecast window opens.

Evidence

  • Soft evidence 7-day antecedent rainfall (IMERG)
    antecedent moisture [scripted]
    BN node: antecedent_rainfall
  • Soft evidence Ensemble-mean exceedance prob (ECMWF P_heavy)
    low-modest [scripted]
    BN node: exceedance_prob
    The ensemble MEAN can look benign for convective floods. Watch the tail.

Act I quiz — build the evidence inventory

Every answer becomes an input to the risk model: evidence → classification → confidence → belief update → decision. Not scored.

Q1. The rainfall outlook for Western Province comes from the ECMWF ensemble (many parallel model runs, updated daily). First, the basics: what is a DETERMINISTIC forecast?

BN purpose: Forecast literacy — why a single number is not enough for risk decisions.

Q2. And what is an ENSEMBLE prediction system (EPS)?

BN purpose: Why forecasts are SOFT evidence in the BN — the spread is the confidence.

Q3. What is the strongest evidence currently available?

On your desk this round: 7-day antecedent rainfall (IMERG) · Ensemble-mean exceedance prob (ECMWF P_heavy)

BN purpose: Determines which BN node receives the highest weight.

Q4. Classify that evidence.

Check the tags on the evidence cards: 7-day antecedent rainfall (IMERG) · Ensemble-mean exceedance prob (ECMWF P_heavy)

BN purpose: Determines confidence weighting — hard = observation (one-hot), soft = probability vector, virtual = likelihood message.

Q5. How reliable is this evidence?

Western Province is a TAIL-risk case — the ensemble mean can look benign while a few members carry the danger; weigh the tail, not the average. Reliability sets how hard the BN leans on it.

BN purpose: Evidence likelihood weighting.

Q6. What hazard condition does this evidence support in Western Province, Rwanda?

Which mechanism fits Western Province — slow river rise, sudden flash runoff, or urban-drainage overload? The mechanism decides which impacts follow.

BN purpose: Updates the hazard node.

Q7. Which impact pathway is most likely here?

Western Rwanda, long-rains. Steep terrain prone to flash floods and landslides; hillside communities exposed. You are the DOC duty analyst as the forecast window opens.

BN purpose: Updates the impact node.

Q8. What evidence would you request next?

Still off the table this round: Ensemble-max / 2-yr RP (pixel p95) · Spatial coverage of heavy-rain pixels · Yesterday's posterior carried forward (DBN, alpha=0.6) · Observed rainfall / river response. Which would most reduce your uncertainty before you decide?

BN purpose: Real DOC operations seek more evidence — watch which cards the Act II rounds reveal.

Q9. Western Rwanda sits on the steep Congo-Nile divide escarpment above Lake Kivu, with cultivated hillsides draining through small, short tributary catchments. Which flood mechanism dominates here?

Think about how fast a small, steep catchment turns rainfall into runoff.

BN purpose: Why short catchment response times make the forecast tail, not the observed river level, the actionable signal here.

Q10. It is late April in the long rains (Itumba) and soils on the western hillsides are already wet from weeks of rain. Why does this antecedent moisture matter for the NEXT heavy-rain forecast?

The first evidence card you see is 7-day antecedent rainfall.

BN purpose: Why the antecedent_rainfall node raises flood risk before the forecast nodes fire.

Q11. Your current risk estimate — before the model runs?

Western Province is a TAIL-risk case — the ensemble mean can look benign while a few members carry the danger; weigh the tail, not the average. No model has spoken yet — commit your own read so Act III can compare it to the engine.

BN purpose: Your pre-BN estimate. Act III compares it against the engine risk indication.

Q12. Recommended DOC status — commit now, before seeing the BN output.

The ensemble mean may look benign but the tail signal is rising ahead of onset. What DOC level do you set, and why? A 'Watch + re-check + note the tail uncertainty' is a strong answer.

BN purpose: Your pre-BN decision. Act III compares it against the CRMA state.

Q13. The risk model you are about to work with was built from expert rules, not calibrated against decades of events. How much should you trust it?

No right answer — your stance is revisited in Act III against what the model actually did, and what history recorded.

BN purpose: Hold this question. Act III returns to it.

DroughtEM-DAT Disaster Events

Monthly Event Frequency (Daily, 2023-04-25 → 2023-05-05, RWA only)

Affected Regions

Admin1 Frequency Choropleth

scroll / pinch to zoom · drag to pan