Weather radars are commonly used to track the development of convective storms due to their high resolution and accuracy. However, the coverage of existing weather radar is very limited, especially in mountainous and ocean areas. Geostationary meteorological satellites can provide near global coverage and near real-time observations, which can compensate for the lack of radar observations. In this paper, a deep learning method was used to estimate the radar composite reflectivity from observations of China’s new-generation geostationary meteorological satellite FY-4A and topographic data. The derived radar reflectivity products from satellite observations can be used over regions without radar coverage. In general, the deep learning model can reproduce the overall position, shape, and intensity of the radar echoes. In addition, evaluation of the reconstruction radar observations indicates that a modified model based on the attention mechanism (Attention U-Net model) has better performance than the traditional U-Net model in terms of all statistics such as the probability of detection (POD), critical success index (CSI), and root-mean-square error (RMSE), and the modified model has stronger capability on reconstructing details and strong echoes.

This study quantifies the characteristics of different satellite sampling errors in the time series of total precipitable water (TPW) derived from Constellation System for Meteorology, Ionosphere, and Climate (COSMIC) radio occultation (RO), Special Sensor Microwave Imager Sounder (SSMIS) and High‐resolution Infrared Radiation Sounder (HIRS) during the overlapping time period of January 2007 to December 2013. Gap‐free data from ERA5 reanalysis of the European Centre for Medium Range Weather Forecasts (ECMWF) are used as reference values. All TPW data are first compared with microwave radiometer (MWR) measurements from Atmospheric Radiation Measurement (ARM) Program. In general, they are consistent, with all their regression coefficients being greater than 0.77. Discrepancies in global TPW time series can be mainly attributed to the inherent sampling errors of these three different satellite remote sensing systems. COSMIC has small sampling errors in higher latitudes. But it has scarce samples in tropical regions which leads to a large sampling error of 3.00 mm in the estimation of global TPW. Sampling in SSMIS is more uniform with mean errors less than 0.5 mm. But the sampling is only over the ocean. Sampling errors in HIRS are larger in tropics and north subtropical areas due to clear sky biased sampling. Moreover, it is significantly correlated with the variability of TPW whereas the sampling error in COSMIC is less influenced by TPW. Sampling errors will be reduced and more consistent global TPW time series will be derived by simply combining the multi‐sensor samplings together.

A near global dataset of homogenized clear-sky 6.5-μm brightness temperatures (BTs) from international geostationary (GEO) weather satellites has recently been generated and validated. In this study, these radiance measurements are used to construct the diurnal variation of upper tropospheric humidity (UTH) and to evaluate these diurnal variations simulated by five reanalysis datasets over the 45° N–45° S region. The features of the diurnal variation described by the new dataset are comparable with previous observational studies that a land–sea contrast in the diurnal variation of UTH is exhibited. Distinct diurnal variations are observed over the deep convective regions where high UTH exists. The evaluation of reanalysis datasets indicates that reanalysis systems still have considerable difficulties in capturing the observed features of the diurnal variation of UTH. All five reanalysis datasets present the largest wet biases in the afternoon when the observed UTH experiences a diurnal minimum. Reanalysis can roughly reproduce the day–night contrast of UTH but with much weaker amplitudes and later peak time over both land and ocean. Comparison of the geographical distribution of the diurnal variation shows that both ERA5 and MERRA-2 could capture the larger diurnal variations over convective regions. However, the diurnal amplitudes are widely underestimated, especially over convective land regions, while the phase biases are relatively larger over open oceans. These results suggest that some deficiencies may exist in convection and cloud parameterization schemes in reanalysis models.

The monthly variation of upper tropospheric water vapor (UTWV) simulated by six reanalysis data sets is evaluated with homogenized water vapor radiance observations from international geostationary (GEO) weather satellites by using a profile‐to‐radiance approach over 45°N to 45°S regions for the period 2015–2017. Results show that reanalysis data sets have an overall good agreement with observations. However, a widespread wet bias is found in all reanalyses and is more dominant in large‐scale subsidence regions. JRA55 has the smallest wet bias while MERRA‐2 exhibits the most humid upper troposphere. The temporal variation of brightness temperatures in response to the warm phase of El Niño‐Southern Oscillation (ENSO) in 2015–2016 indicates that the UTWV field is regulated by both ascending and descending branches of the large‐scale circulation. All six reanalyses roughly capture the temporal variation of UTWV in the developing and decay year of this ENSO event. However, they tend to overestimate the eastward propagation of high UTWV in the developing year, especially MERRA‐2. The UTWV gradient over the tropical Pacific in the decay year is underestimated, with a dry bias over the convective western Pacific and a wet bias over the eastern Pacific in reanalysis data sets. These results may provide a useful tool for the climate modeling community for identifying and solving problems associated with UTWV simulation.

The distribution of tropospheric moisture in the environment is highly associated with storm development. Therefore, it is important to evaluate the uncertainty of moisture fields from numerical weather prediction (NWP) models for better understanding and enhancing storm prediction. With water vapor absorption band radiance measurements from the advanced imagers onboard the new generation of geostationary weather satellites, it is possible to quantitatively evaluate the environmental moisture fields from NWP models. Three NWP models—Global Forecast System (GFS), Unified Model (UM), Weather Research and Forecasting (WRF)—are evaluated with brightness temperature (BT) measurements from the three moisture channels of Advanced Himawari Imager (AHI) onboard the Himawari-8 satellite for Typhoon Linfa (2015) case. It is found that the three NWP models have similar performance for lower tropospheric moisture, and GFS has a smaller bias for middle tropospheric moisture. Besides, there is a close relationship between moisture forecasts in the environment and the tropical cyclone (TC) track forecasts in GFS, while regional WRF does not show this pattern. When the infrared and microwave sounder radiance measurements from polar orbit satellite are assimilated in regional WRF, it is clearly shown that the environment moisture fields are improved compared with that with only conventional data are assimilated.

Although atmospheric CO₂ is a trace gas, it has seasonal variations and has increased over the last decade. Its seasonal variation and increase have substantial radiative effects on hyperspectral infrared (IR) radiance calculations in both longwave (LW) and shortwave (SW) CO₂ absorption spectral regions that are widely used for weather and climate applications. The effects depend on the spectral coverage and spectral resolution. The radiative effect caused by the increase of CO₂ has been calculated to be greater than 0.5 K within 5 years, whereas a radiative effect of 0.1–0.5 K is introduced by the seasonal variation in some CO₂ absorption spectral regions. It is important to take into account the increasing trend and seasonal variation of CO₂ in retrieving the atmospheric temperature profile from hyperspectral IR radiances and in the radiance assimilation in numerical weather prediction (NWP) models. The simulation further indicates that it is very difficult to separate atmospheric temperature and CO₂ information from hyperspectral IR sounder radiances because the atmospheric temperature signal is much stronger than that of CO₂ in the CO₂ absorption IR spectral regions.