• African scientists are leveraging low-cost 3D-printed weather stations to address the sharp decline in functioning ground-based observation networks, which are critical for accurate weather monitoring and forecasting.
• The innovative 3D-Printed Automatic Weather Station (3D-PAWS) model allows countries to locally build, repair, and maintain stations for a fraction of traditional costs, significantly improving data availability and sustainability.
• Enhanced ground data, combined with satellite imagery, is strengthening early warning systems for extreme weather, helping governments and aid organizations better anticipate droughts, floods, and food security crises.

Satellites give us an out-of-this-world vantage point to track weather conditions in every corner of the globe.

But satellite imagery alone can only paint part of the picture. Scientists also rely on Earth-bound weather stations to calibrate satellite information and accurately track and monitor rainfall.

By combining satellite imagery with ground-based weather station data, scientists with the Famine Early Warning Systems Network (FEWS NET), their counterparts at national meteorological centers, and climate scientists worldwide are able to better monitor and predict weather patterns, forecast agricultural production outcomes, and prepare for potential food emergencies.

“We’ve gotten really good at monitoring extremes and making predictions,” Chris Funk, Director of the UC Santa Barbara Climate Hazards Center, said. “We are very good at rapidly ‘stacking’ the information from disparate sources, like satellites and rain gauges, and this allows us to identify extreme droughts almost in real-time, with a high degree of confidence.”

In recent years, the number of functioning ground-based weather stations in regions like the eastern Horn of Africa and Southern Africa has declined, resulting in a reporting crisis that has made it difficult for scientists to produce rainfall estimates that are fundamental to monitoring our changing climate. 

Reporting crises occur when a considerable number of available weather stations “drop off” and stop recording or transmitting data as expected. Weather stations can drop off for a variety of reasons, including dead batteries, glitches in transmission systems, inadequate network connectivity, or other structural issues that keep stations from collecting rainfall measurements in the first place.

In Africa, the number of available stations has declined from about 3,300 in 1981 to less than 800 in 2023.

“We’ve been running into headwinds with the weather stations that were commercially bought in markets,” Absae Sedah, Assistant Director of Meteorological Services with the Kenya Meteorological Department, said. “Getting replacement parts for these stations has become a very big challenge for us.”

According to Sedah, the Kenyan government has spent nearly $25,000 on each weather station it has procured from a variety of international vendors. 

Because of the high costs of repairs, and due to the logistical constraints associated with calibrating stations that come from different manufacturers, Sedah said only 170 of the Kenya Meteorological Department’s 400 weather stations are currently functioning.

Paul Kucera and Martin Steinson with the University Corporation for Atmospheric Research (UCAR) said, for decades, international meteorological organizations have been aware of the structural and financial challenges with commercial weather stations.

In 2013, UCAR scientist Kelly Sponberg came to Kucera and Steinson with an idea for a sustainable solution.

“This idea of 3D printing really became affordable in the early 2010s as the technology developed,” Kucera explained. “This is kind of how the idea came to set up fabrication labs in countries and teach people how to make their own 3D-printed weather stations.”

What resulted was a prototype for the first 3D-Printed Automatic Weather Station (3D-PAWS), representing a potential paradigm shift in the capacity for local meteorological organizations to fabricate and maintain weather stations independently and at low costs.

3D-PAWS surface weather stations can be manufactured in just one week for between $400–$600 USD using locally sourced materials, microsensor technology, low-cost microcontrollers, and a 3D printer.

“It started to catch on as a technology that could be used for filling the gaps that existed with the traditional high cost weather stations,” Kucera said. 

According to Kucera and Steinson, sustainability and adaptability were key components of the 3D-PAWS development process. The 3D-printed weather stations were designed to use locally available materials where possible, such as PVC pipe and fittings to construct the frame.

After years of fine tuning the 3D-PAWS model and performing validation exercises to ensure accurate rainfall measurements, in March, the White House announced USAID/FEWS NET’s $10 million commitment to help 10 African governments learn how to install, use, and maintain 3D-printed weather stations to bolster climate, weather, and acute food insecurity forecasts. 

Following a training workshop in June 2023, Kenya became the first FEWS NET country to adopt 3D-PAWS. By the end of September, the Kenya Meteorological Department successfully fabricated and installed 24 weather stations across the country.

“[The trainees with the Kenya Meteorological Department] are excited and getting interested in having something that they can call their own and having something that they can manage independently,” Sedah said. “That's very much key for us.”

Kenya’s new 3D-printed weather stations are transmitting data every 15 minutes with observations covering rainfall, temperature, pressure, and relative humidity. According to FEWS NET Regional Scientist for East Africa Gideon Galu, these observations allow scientists to assess real-time weather conditions and issue weather advisories in areas where extreme climate events frequently occur.

While there are major benefits to increasing the use of 3D-printed weather stations, scientists will need to deal with challenges related to maintenance, as well as communicating and interpreting the information they collect.

“All weather stations can experience problems, which can be hard to detect, and reaching the stations for maintenance can be expensive,” Funk said. “It can also be challenging to effectively communicate appropriate advice at local scales, so it will be important to provide ongoing support for maintenance, analysis, and last-mile communication.”

As FEWS NET scientists continue to validate data collected by Kenya’s new 3D-printed weather stations during the ongoing rainy season, a second training workshop is set for Zimbabwe.

“Next up we’re taking the 3D-PAWS project to Zimbabwe. Our kickoff meeting in Harare with the Zimbabwe Meteorological Services Department went wonderfully,” Funk said. “So far, the results of our training in Kenya have been really encouraging. We’re excited to see a very low margin of error with our first new weather stations.

“Having this data validated and available, especially as we’re in the middle of a very strong El Niño event and positive Indian Ocean Dipole, is going to really help in tracking extreme rains so we can anticipate impacts on agricultural production and food security. The efforts of Gideon Galu and Tamuka Magadzire, our regional scientists in Kenya and Botswana, combined with the deep expertise of UCAR’s International Capacity Development Program, and the sterling efforts of the Kenya Meteorological Department, have made a lot of progress over a few short months.”

Sign up to receive FEWS NET reports and updates in your inbox.