Reply on RC3

Abstract

<strong class="journal-contentHeaderColor">Abstract.</strong> Four upper tropospheric humidity (UTH) datasets derived from satellite sounders are evaluated to assess their consistency as part of the activities for the Global Energy and Water Exchanges (GEWEX) water vapor assessment project. The datasets include UTH computed from brightness temperature measurements of the <span class="inline-formula">183.31±1</span> GHz channel of the Special Sensor Microwave – Humidity (SSM/T-2), Advanced Microwave Sounding Unit-B (AMSU-B), and Microwave Humidity Sounder (MHS) and from channel 12 of the High-resolution Infrared Radiation Sounder (HIRS). The four datasets are generally consistent in the interannual temporal and spatial variability of the tropics. Large positive anomalies peaked over the central equatorial Pacific region during El Niño events in the same phase with the increase of sea surface temperature (SST). Conversely, large negative anomalies were obtained during El Niño events when the tropical-domain average is taken. The weakened ascending branch of the Pacific Walker circulation in the western Pacific and the enhanced descending branches of the local Hadley circulation along the Pacific subtropics largely contributed to widespread drying areas and thus negative anomalies in the upper troposphere during El Niño events as shown in all four datasets. During a major El Niño event, UTH had higher correlations with the coincident precipitation (0.60 to 0.75) and with 200 hPa velocity potential (<span class="inline-formula">−0.42</span> to <span class="inline-formula">−0.64</span>) than with SST (0.37 to 0.49). Due to differences in retrieval definitions and gridding procedures, there can be a difference of 3 %–5 % UTH between datasets on average, and larger magnitudes of anomaly values are usually observed in spatial maps of microwave UTH data. Nevertheless, the tropical-domain averaged anomalies of the datasets are close to each other with their differences being mostly less than 0.5 %, and more importantly the phases of the time series are generally consistent for variability studies.