impactoutcome™ETHIOPIA
ImpactOutcome Global Africa Ethiopia Dire Dawa Flood early warning

Dire Dawa · Flood early warning

Earlier warnings. Safer neighbourhoods.

Flash floods from the Dechatu River can reach Dire Dawa's neighbourhoods with little notice. This proposal connects the weather stations, river gauges, forecasts and restored landscapes already in place into one early-warning service, so the city can warn residents sooner and act together.

PROPOSED PROJECT · FUNDING BEING SOUGHT

Watch

The idea in 60 seconds.

This page and video describe a proposed project contemplated for delivery beginning in 2027. It is subject to funding, partner approvals and the outcome of an open procurement; no technology supplier has been selected. Scope, locations and timing may change. Imagery is illustrative and does not depict existing facilities, equipment or people. Nothing here is a commitment by any institution named or shown.

The challenge

Heavy rain upstream, homes downstream.

165 km²Dechatu River catchment above and around the city
580K+urban and peri-urban residents who stand to benefit
24 moto design, test and hand over a city-owned service

Intense rainfall, eroded hillsides and fast urban growth mean water arrives quickly. Women, young people and low-income households in flood-prone kebeles are hit hardest.

What's already there

The pieces exist. The link is missing.

920+
hectares of restored hillsides, agroforestry and river buffers, reported by earlier catchment work
4 + 3
automatic weather stations and river-level monitors reported in the catchment
The gap
no end-to-end service that turns those readings into timely warnings and coordinated action

Station condition, ownership and data access will be verified on the ground before anything is built on them.

How it works

AI does the watching. People make the call.

Illustration: a riverside weather and river-level station01 SENSE

Watch the river and the sky

Weather stations, river gauges, official forecasts, maps and community reports, checked for quality.

Illustration: city staff reviewing a flood-risk map02 ASSESS

See risk rising early

AI flags changing flood risk with its sources, timestamps and uncertainty shown, never as a black box.

Illustration: city officials coordinating a response03 APPROVE

Officials decide

Authorised officials approve every public warning. Each decision is logged and reviewed after events and drills.

Illustration: a resident receiving a warning on a phone04 WARN & ACT

Reach every household

Radio, SMS, community networks and digital channels, in local languages, with a manual fallback when the network is down.

Six helpers, tightly bounded

Each AI agent has one job.

Agents support trained staff. They do not issue warnings or make emergency decisions on their own.

Data check

Checks weather-station and river-level readings for gaps and errors before anyone relies on them.

Risk watch

Combines readings, official forecasts and maps to flag when flood risk is rising, and how sure it is.

Community reports

Organises reports from residents and kebele contacts so local knowledge reaches the right desk.

Nature sites

Keeps track of restored hillsides, agroforestry plots and river buffers that help hold water back.

Situation updates

Drafts short briefings for city departments, so everyone works from the same picture.

Resident messages

Drafts clear warnings in local languages for officials to approve, edit or reject.

Nature as infrastructure

Protecting the landscapes that hold the water back.

Illustration: restored green hillsides above a river valley

Restored hillsides, agroforestry and river-buffer areas slow runoff, reduce erosion and protect soil and water. The project adds a simple monitoring and learning tool so the city can see which sites are working, which need care, and how they relate to flood risk downstream. No new planting or construction is proposed.

Principles

Built for trust, built to last.

People in charge

Public warnings and emergency decisions stay with designated government officials.

Everyone included

Women, young people, older people and people with disabilities help design and test the warnings.

Careful with data

Only the data needed, with clear rules for consent, access and retention.

City-owned

The city owns the operating procedures. Local staff are trained to run the system, with a costed plan to keep it going.

The path

Twenty-four months, four stages.

Map and verifyCheck existing stations, data and restored sites; choose pilot kebeles with communities.
Design and buildAgree the warning workflow and approval rules; build and secure the system.
Test and rehearseRun simulations with city staff and residents; test messages for clarity and reach.
Hand over and shareTrain local teams, hand over, and publish lessons for other Ethiopian cities.

Partners

A city service, led locally.

Haramaya UniversityLead applicant · science, evaluation and learning
Dire Dawa Administration Environmental Protection AuthorityCity partner · owns the warning process

Specialist technical services will be chosen through open, competitive procurement. No technology supplier has been selected.

Beyond Dire Dawa

A model other cities can adapt.

The aim is for the Dechatu catchment to become a national learning hub: tested procedures, training materials and a modular design that other flood-exposed cities and catchments in Ethiopia can adapt to their own risks, languages and institutions.

This page and video describe a proposed project contemplated for delivery beginning in 2027. It is subject to funding, partner approvals and the outcome of an open procurement; no technology supplier has been selected. Scope, locations and timing may change. Imagery is illustrative and does not depict existing facilities, equipment or people. Nothing here is a commitment by any institution named or shown.