Picture: SpaceX Satellite. Credit: SpaceX (Pexels).
The absence of ground observations in many parts of the world highlights the importance of satellite products for capturing precipitation.
In this study, CLEX researchers aimed to evaluate the effect of different sources of data and the uncertainties in satellite data, by comparing the data with a ground-based radar product using both location-based and storm-based approaches. The study focused on the eastern United States (land-only) during hurricane days that occurred in 2016–2018.
The results showed the satellite data had better agreement in terms of the average precipitation intensity and area. However, the satellite observations tended to show storms with smaller areas compared to the ground-based observations, possibly due to the effect of light precipitation not being detected properly.
If this light precipitation (less than 1 mm/h) is removed in an object-based approach, hurricane objects in the satellite observations tend to be larger, which might be related to different viewing angles of sensors contributing to these observational datasets.
Precipitation estimates have smaller areas with higher average intensity in the satellite observations, which is probably related to the effect of a morphing technique that leads to the homogenisation of varying rainstorm characteristics.
The researchers were also able to distinguish what caused the reduced precision of satellite observations that led to a growth in precipitation area and a downward trend in average precipitation intensity compared to ground-based observations. They were also able to show, which satellite sensors were best at capturing observed characteristics, the average intensity of precipitation events, and precipitation area.
Overall, storm characteristics viewed from satellites are similar enough to ground observations to recommend their use more broadly.
- Paper: Ayat, Hooman, Jason P. Evans, and Ali Behrangi. “How Do Different Sensors Impact IMERG Precipitation Estimates during Hurricane Days?” Remote Sensing of Environment 259 (June 15, 2021): 112417. https://doi.org/10.1016/j.rse.2021.112417.