Climatological evaluation of satellite and reanalysis precipitation products across temporal scales and rainfall extremes over the Guder Sub-basin, Ethiopia
This study evaluates six satellite and reanalysis precipitation products against gauge data in Ethiopia's Guder Sub-basin, revealing that while performance varies by temporal scale and rainfall intensity, CHIRPS, MSWEP, and IMERG emerge as the most reliable datasets for hydrological and climate applications.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Rain is the lifeblood of the Ethiopian highlands, feeding the rivers that carve through the landscape and filling the fields where farmers grow their crops. Yet, in many parts of this rugged terrain, the ground truth of how much rain actually falls remains a mystery. The network of weather stations that measures rainfall is sparse, with large gaps between instruments, leaving scientists to guess the patterns of storms across valleys and peaks. To fill these gaps, researchers have turned to the sky, using satellites and computer models to estimate precipitation from space. These digital estimates are now essential tools for planning agriculture, managing water resources, and preparing for floods or droughts. But a critical question remains: how accurate are these sky-based guesses when compared to the reality of the ground, especially in a place as complex as the Guder Sub-basin, a steep, mountainous tributary of the mighty Blue Nile?
A team of researchers set out to answer this question by putting six of the most popular satellite and computer-model rainfall products to the test. They focused their study on the Guder Sub-basin in Ethiopia, a region where the land rises sharply from 1,500 to 3,000 meters, creating a patchwork of microclimates that are difficult to capture from orbit. The scientists gathered fifteen years of daily rainfall records from seven reliable weather stations scattered across the basin. They then compared these ground measurements against data from six different sources: CHIRPS, IMERG, TAMSAT, MSWEP, PERSIANN-CDR, and ERA5. Each of these products uses a different method to estimate rain; some blend satellite images with ground data, while others rely entirely on computer simulations of the atmosphere. The team checked how well each product performed not just on average, but across different timeframes—day by day, month by month, and through the distinct wet and dry seasons—and even when it came to the most extreme weather events.
The investigation revealed that no single product was perfect for every situation, but some stood out clearly depending on the task. When looking at daily rainfall, the product known as MSWEP proved to be the most reliable. It matched the ground observations with the highest accuracy, capturing the daily ups and downs of rain better than the others. However, as the researchers looked at longer periods, the story shifted. When the data was averaged out over a month, two other products, CHIRPS and IMERG, emerged as the most trustworthy, showing a very strong agreement with what the weather stations actually recorded. This suggests that while one tool might be best for seeing the rain of a single day, another might be better for understanding the total water available over a season.
The performance of these tools also changed dramatically with the seasons. During the small rainy season in the spring and the dry season in the winter, most of the products did a decent job of tracking the weather. But when the main rainy season arrived, bringing the intense, heavy monsoonal storms from June to September, the accuracy of all the satellite and model estimates dropped. The complex, powerful storms of this season proved difficult for any of the digital tools to capture precisely. One product, ERA5, consistently struggled across all timeframes and seasons, often overestimating the amount of rain by a wide margin, making it the least reliable choice for this specific region.
The researchers also examined how well these products could predict the extremes of the weather, such as long dry spells, consecutive wet days, and the most violent downpours that cause flooding. Here, the results were mixed. For tracking dry spells, PERSIANN-CDR performed best, while MSWEP was the most accurate for tracking wet spells. When it came to moderate rain, CHIRPS was a strong performer, but for the heaviest, most dangerous rainfall events, IMERG showed the best ability to detect them. However, even the best tools struggled with the most extreme outliers, such as the rare, massive storms that define flood risks. The study concluded that while these satellite and model products are invaluable for a region with so few ground stations, they are not interchangeable. For anyone using this data to manage water or plan for climate risks in the Guder Sub-basin, the choice of tool must be specific to the need: MSWEP for daily details, CHIRPS or IMERG for monthly totals, and a careful selection for extreme events, always remembering that local conditions require local verification.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.