Station Log
What the instruments on Kenyan farms actually recorded — rainfall, wind and the hours you can spray. Measurements, not forecasts.
How cold it gets overnight in the Kenyan highlands
The lowest reading across seven stations in a month was 3.7 °C, well short of frost. Soil probes averaged 17.7 °C over the same weeks, and no published Kenyan source gives a minimum series for a highland station.
Was this month's rainfall enough for your crop?
A gauge total is only the first number. Our stations logged 0.4 mm to 68.9 mm over the same 30 days against 45 mm of reference demand, and Kisumu's 134 percent April in 2021 still lost the late planting.
Running a weather station with no mains power
Solar answers the power question. The harder one is how the reading leaves the farm: test the Safaricom signal at the mast site first, because a station on a dead patch is an expensive rain gauge nobody reads.
Why the nearest weather station does not describe your farm
Four of our stations recorded between 0.4 mm and 65.7 mm of rain over the same four weeks, on farms not far apart by Kenyan standards. Temperature travels well across that distance; rainfall does not.
What a farm weather station costs in Kenya
A connected station runs KES 30,000 before taxes, a soil sensor from KES 7,000, and SIM data about one US dollar a month. But if rainfall total is your only question, a manual gauge answers it for almost nothing.
Where to find rainfall records for your part of Kenya
Over the same month, four stations recorded totals from 0.4mm to 104.4mm, so a five-kilometre satellite cell will not settle a spray decision. Match the source to the question you are asking this week.
What 0.2 mm of rain actually means on your farm
In Kenyan records 0.2mm is the line between a dry day and a wet one, not a bucket size. Measured tips run from 0.221mm to 0.367mm, and every gauge undercounts once rain passes about 50mm an hour.
Rainfall intensity: how fast the rain fell, not just how much
Across eight Kenyan stations last month, only 1.1 percent of ten-minute readings carried any rain at all, so the water came in short bursts. Your gauge already logs every tip; the rate is in there, waiting to be worked out.
How long leaves stay wet matters more for disease than how much rain fell
Many fungi need about nine unbroken hours of free water on the leaf. Our five stations averaged 6.2 wet hours a day for a month, while a coffee study across six Kenyan counties put the rust line at ten hours with humidity above 75 percent.
Weather station siting: how much clear ground each sensor needs
Across 11 stations in Kenya over one month, rainfall totals ran from 0 to 157.8 millimetres and temperature from 3.9 to 31.3 degrees. That spread is the sum of eleven different siting compromises, made before any reading was logged.
Weather station siting problems: usually an afternoon's work to fix
A mast that has reported for two seasons can still be reading a fence, a wall or an afternoon shadow. Three responses are on the table before the short rains, and they cost very differently.
Rain gauge errors: undercatch, spacing and how to fix them
A rain gauge looks like the simplest instrument on the farm: a funnel, a bucket, a number. The evidence says the number moves for reasons that have nothing to do with how much rain actually fell.