On a monthly timescale, IMERG generally agrees well with the rain gauges and is able to capture the distinct characteristics of the seasonal cycle of rainfall. However, the evaluation of the sub-daily timescale reveals three salient aspects: (1) IMERG is prone to false alarms that account for almost a quarter of total rainfall; (2) IMERG overestimates the rainfall amount for frequently occurring weak convective events, while that of relatively rare, but strong mesoscale convective systems is underestimated, resulting in an error compensation; and (3), the skill decreases during the local little dry season in July and August, which is known to exhibit enhanced low-level cloudiness. These findings, in turn, are related to: (1) a general oversensitivity in the sensors towards clouds with low optical thickness; (2) a strong negative bias for high intensities, irrespective of the overpassing IMERG sensor; and (3), a large fraction of missed events linked with “warm rain”. The latter is pronounced whenever IR observation is included, while direct passive microwave overpasses mitigate this problem.
Despite the aforementioned deficiencies on smaller time scales, this work stresses the overall significance of IMERG for data-sparse regions such as the West African Guinea Coast. However, rain gauge mesonets can help to identify region-, season-, and rainfall system-dependent differences in the performance of IMERG. Their data can then also be used for bias-correcting GPM IMERG rainfall estimates.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner