How NGOs watch the weather at every field site
Simple weather monitoring for NGO field sites: one forecast note a day for every programme location, with heavy rain, wind and heat flagged for the teams on the ground.
Part of the guide Workflow automation
The simplest weather monitoring for NGO field sites is a note that arrives every morning with the forecast for each programme location and flags only what changes plans: heavy rain that will cut a road to a village, strong wind before a distribution under tents, heat that makes an afternoon outreach session unsafe. Country offices that run dozens of sites cannot check a forecast for each one by hand, but they can read one list that puts the three sites that need attention at the top.
This article is for programme managers, logistics officers and security focal points at humanitarian organisations, development charities and community groups with work spread across rural areas. It covers what to forecast and for how long, how to set thresholds, a made-up example across three countries, the limits of forecasts, and the AI employee that sends the note.
What to forecast, and for how far ahead
Field decisions sit on two timescales. Today and tomorrow decide whether a team travels, whether a clinic opens outdoors and whether a delivery goes ahead. The next week to ten days decides when to move stock before roads soften, or when to bring a distribution forward.
The Weather forecast block covers both: up to 16 days ahead, from national weather services, with temperature, rain, wind and UV. Given a list of sites, it adds to each one today's weather, the rain expected in total and on how many days, the first rainy day, the strongest wind, the highest temperature and a one-line summary ready for an email.
- Rain matters most for access: dirt roads, river crossings, flooding of low ground.
- Wind matters for tents, temporary shelters and anything stored outside.
- Heat and UV matter for outreach, queues at distributions and the people waiting in them.
Setting thresholds that mean something
Forecasts arrive as numbers; teams need a yes or no. Agree with each country team what counts as a problem for them, because 20 mm of rain is a nuisance on a paved road and a closure on a laterite track. Typical starting points, over the next three days, are 20 mm or more of rain in all, wind of 50 km/h or more, or a high of 38 degrees, then adjusted after a month of use.
A made-up example across three countries
Imagine an invented health charity with 28 sites: mobile clinic stops in Kenya's Turkana region, school feeding points in northern Bangladesh and community water projects in the Peruvian highlands. Its regional office keeps the sites in a spreadsheet with their positions.
At 06:00 Nairobi time, in this made-up case, the note lists all 28 sites but opens with three flagged ones: two Bangladesh feeding points with about 45 mm of rain expected tomorrow, where the logistics officer moves the rice delivery forward a day; and one Peru site with strong wind forecast, where the water team ties down the storage tanks. The other 25 sites appear below in one line each, so nobody wonders whether they were checked.
Building the daily note
In xMap Flows the note is an AI employee made of a few blocks:
- Schedule: every day at 06:00 in the regional office's time zone.
- My data: the table of sites, each with a name, country team, position and the email of its focal point. A sheet in Google Sheets works too, so field staff can add sites themselves.
- Weather forecast: each incoming place, the next 3 days, with rain, wind and temperature.
- Calculated field: a flag that says yes when rain, wind or temperature passes your agreed limits.
- Sort and limit: flagged sites first, then by rain expected.
- Email: one note to the regional team with every site, or one email per site to its focal point.
Up to 500 sites can be checked in one run, and sites close together share a forecast. For a quick look at a wider area, such as a district after a storm warning, set the block to the Area of interest and read the forecast day by day. Ready flows that start from the same idea are heavy rain at our camps and settlements, rain alert for my city and morning site briefing: weather and road closures, which adds road closures from Traffic incidents where online maps cover them.
Mistakes to avoid
- One threshold for every country. Highland rain and delta rain behave differently. Let each team set its own limits.
- Positions that are only approximate. A site placed at the district capital gets the capital's forecast. Use the real position, taken from a phone if needed.
- A note too long to read. If the whole list runs past one screen, send each focal point their own sites and the regional office the flagged ones only.
- Treating it as an early warning system. It is a planning aid built on public forecasts, not a replacement for official warnings or your security procedures.
The non-profits and humanitarian page lists more flows for programme teams, from food bank gaps to relief news on a map. How field teams hand this kind of daily checking to software is covered in the guide to an AI employee for field operations. To discuss your own sites, book a call.
Questions
Does it work in remote rural areas?
Yes. The forecast is worked out for any position on land, so a village or a clinic stop gets its own forecast even far from a city.
Can field staff receive it on their phones?
Yes, by email, or in a Slack channel they already use. Each site's focal point can get only their own sites.
How far ahead can it look?
Up to 16 days. Most teams use 3 days for decisions and 7 to 10 days for preparing stock and travel.
Can it warn about floods directly?
It forecasts rain, not river levels. Heavy rain at or upstream of a site is the signal to watch, together with local knowledge and official warnings.