Rethinking Tropical Weather Forecasting - a Moonshot Project

Liisa Petrykowska is engineering better decision making for farmers in the tropics. Most subSaharan countries do not have access to weather reports and forecasts, and the weather patterns have been erratic and harder to predict over the last few decades.

In her Solve For X talk Tropical Weather Forecasting Petrykowska discusses her weather forecasting system, the underlying theories behind the method, and a vision for the next few decades.  Many farmers in the target regions of the study (Liisa’s company ignitia has offices in Sweden, Nigeria and Ghana) are responsible for providing labor and food for their extended family, but the traditional ways of farming are becoming obsolete.


https://www.solveforx.com/moonshot/5735240957952000

Petrykowska explains how the tropic band of land around the equator has different weather patterns and different physical characteristics from the rest of the world. Around two billion people live in this band and eighty percent of them depend on agriculture for their living, mostly small scale farmers.

Understanding the physics behind the convective weather patterns of the region and modifying the resolution used to predict thunderstorms is the largest contributor to ignitia’s success. Fifteen years and ten developers created the algorithms to simulate tropical weather. Their result is an eighty four percent average accuracy of forecasts over three seasons vs the thirty nine percent accuracy of the three leading weather prediction bodies.

Crop selection, land prep, labor planning, applying fertilizer, spraying chemicals and harvesting can all be optimized when a farmer has a good idea of what the weather will do on a given day. Over the last three seasons that Liisa has worked with the farmers in the target regions an increase of fifty to eighty percent yield has been recorded.

The entire process of taking the satellite data and filtering it through algorithms before sending it to the farmers is outlined. After some trial and error ignitia learned that the farmer cares about one thing – is it going to rain? Temperature, humidity, and wind are data points that might help but do not give farmers enough information to act.

This method of weather prediction is amazing, and if the gains made in the pilot areas could be transferred throughout the region incredible change could happen. The food security problems throughout Africa could be eased, and the eighty percent of people relying on farming could have more resilience in their lives. Petrykowska is currently tackling the challenges of scaling up the current successes, contracting telecom operators and recruiting more employees to handle the influx of one thousand new customers every day.


https://www.solveforx.com/moonshot/5735240957952000

(This talk starts around the 5:00 minute mark after the introduction is completed)