Cloud‐Rain Vertical Inconsistency Increases IMERG Precipitation Uncertainty

Abstract IMERG accuracy is limited by the vertical inconsistency between satellite‐observed cloud‐top information and the true vertical structure of precipitation. Using FY‐3G Precipitation Measurement Radar (PMR) observations as reference and FY‐4B cloud‐top parameters, we investigate the vertical‐structure sources of IMERG retrieval errors during the 2024 pre‐summer rainy season in South China. IMERG intensity biases vary markedly with classification framework: when classified by cloud‐top height (CTH), IMERG overestimates precipitation associated with high clouds, whereas classification by storm‐top height (STH) reveals significant underestimation. IMERG detection accuracy increases rapidly as STH exceeds 7.5 km and remains above 0.8. Three cloud‐rain vertical inconsistency regimes are identified: severe missed warm‐cloud precipitation (STH ≈ CTH < 5 km), overestimated precipitation area in anvil clouds (CTH ≫ STH), and underestimated precipitation intensity in convective cores (STH ≈ CTH > 12 km). These results suggest cloud‐rain height difference as a useful correction factor for future multisource precipitation retrieval algorithms.