The three-dimensional (3D) atmospheric temperature structure in typhoon regions plays a critical role in typhoon physics research and numerical prediction, yet direct observational methods have long been lacking. Microwave remote sensing, with its unique cloud-penetrating capability has become a key approach for obtaining thermal structures within typhoons. However, the insufficient temporal resolution of traditional polar-orbiting satellite observations severely limits the application timeliness of microwave data in monitoring rapidly evolving weather systems like typhoons. With the operational deployment of China’s new-generation dawn-dusk orbit satellites, a tri-orbit collaborative observation system comprising FY-3E (dawn), MetOp-C (morning), and NOAA-20 (afternoon) satellites achieves six daily high-frequency observations of typhoons. This study integrates microwave temperature sounder (MWTS-III/AMSU-A/ATMS) data from three orbital platforms and establishes a real-time 3D temperature field reconstruction system with six daily outputs, based on our previously developed typhoon-specific retrieval algorithm. Analyses demonstrate that the retrieved temperatures effectively capture 3D structural features and spatiotemporal variations of typhoon temperature fields. Validation against best-track data reveals a strong correlation (>0.9) between retrieved warm-core intensity and typhoon central pressure, showing 6-9 hours earlier detection capability of intensity changes compared to GDAS-FNL and ERA5 reanalysis datasets. Case studies from the 2023 Northwest Pacific typhoon season confirm that our product maintains more stable temporal correlations with typhoon intensity (>0.9) than GDAS-FNL (0.85) and ERA5 (0.72). These results prove the reliability of the tri-orbit satellite-based temperature retrievals for analyzing typhoon thermal evolution.

Microwave observations have become the primary means of acquiring internal information on typhoons because microwave radiation can penetrate clouds and precipitation. By exploiting the strong relationship between the observed microwave brightness and atmospheric temperature, it is possible to derive a 3-D temperature structure within typhoons. However, these retrievals are limited by the consistent overestimation of warm cores and excessive intensity of cold centers at lower levels. Through the optimization of clear-sky and cloudy data classification and the incorporation of a latitude zone regression algorithm, this study successfully established an improved algorithm for the retrieval of typhoon inner atmospheric temperature using Microwave Temperature Sounder-III (MWTS-III) on China’s FY-3E, the first civil dawn-dusk polar orbit satellite. The analysis of the actual retrieved results for typhoons Bolaven and Doksuri at different stages demonstrates that the new algorithm effectively reduces errors in the low cold center and provides a more reasonable position for the warm core. Further analysis indicated that the accurate separation of quasi-clear-sky and cloud data is crucial for adjusting the warm-core position and mitigating errors in the low-layer temperature bias. In addition, fitting within latitudinal bands helps minimize the overall warm-core bias resulting from latitudinal temperature gradients, allowing for better characterization of real-time changes in typhoon intensity through temperature retrieval. The optimized 3-D air temperature retrieval product for typhoon regions holds great promise for advancing research on typhoon mechanisms and forecasting.

Unified gridded satellite observation data is important for continuous observation of weather systems. However, quality of the grid datasets is largely impacted by the data gaps. This article proposes a contour-aware data reconstruction network (CARNet) using a self-supervised frame to reconstruct the gridded Advanced Microwave Sounding Unit A (AMSU-A) brightness temperature (BT) data. In CARNet, two specified training masks are designed to cope with the spatial characteristic of the data gaps. Moreover, a contour-aware network is developed to integratea prioriknowledge, the continental contour information into the CARNet. In this study, the AMSU-A BT data of 2018-2020, and the daily averaged sea surface temperature data of 2018 are used. Since AMSU-A provides global observation twice per day, for example, a total of 730 samples of AMSU-A data of 2019 are used for training, validation, and testing. On the testing data with over 50% grids being gaps, the CARNet outperforms the other conventional reconstruction methods by achieving the mean absolute error (MAE) of 2.68 K, 2.64 K, and the mean absolute percentage error (MAPE) of 1.32%, 1.30% in the 2019 and 2020 data, respectively. Moreover, the reconstructed data can exhibit more clear day-night trend, seasonal trend, more complete hurricane structures, and homogeneous distribution in regions like the Amazon basin. The results indicate that the proposed CARNet can provide better reconstruction data for the weather and climate research applications.

As a new generation of microwave temperature sensor independently developed by China, the Microwave Temperature Sounder-III (MWTS-Ⅲ) onboard the polar-orbiting satellite FY-3E has 17 channels that provide temperature information from multiple atmospheric layers, making it an indispensable instrument for meteorological research. However, due to the limb effect caused by its cross-track scanning mode, it is impossible to directly observe weather changes using the MWTS-Ⅲ brightness temperature data. Currently, the limb correction algorithm is an important method to eliminate the limb effect. However, it is difficult to apply existing limb correction methods with fixed parameters due to the newly-added channels of MWTS-Ⅲ. The results of our study show that while the old algorithm can effectively correct the limb effect in most channels of MWTS-Ⅲ, there are still some limb effects in the corrected results of channels 5-7. On the other hand, the new algorithm can significantly improve the shortcomings of the old method and more accurately eliminate limb effects. In addition, by comparing the limb correction results under different training data, we found that the correction results of the new algorithm have a certain dependence on the training data, and using the same month in the previous year can achieve better correction results. Finally, it has been proved to be applicable to all channels of MWTS-Ⅱ onboard FY-3D. Specially, the new algorithm can improve the shortcomings of the old algorithm in correcting the limb effect for MWTS-Ⅱ channel 7, including the lower average brightness temperature at the orbit edge, discontinuous spatial distribution of brightness temperature, and the existence of abnormally low values.

Accurate intensity estimation of tropical cyclone (TC) is crucial to the success of disaster warning and management. Extended from previously designed multi-feature distribution learning (MFDL) network and the related routine, this study proposes a multi-view feature fusion (MVFF) network to optimally exploit the synthetic temperature anomalies of TCs sampled in three directions, the along-track, cross-track and pressure-level. The MVFF contains three parallel branches developed from the single-shot detector (SSD) to extract features from the three direction anomalies, respectively. In order to effectively mingle the features extracted from the three branches, MVFF adds a feature effective fusion (FEF) module into each SSD. In this study, 500 TCs occurred in the North Atlantic (NA) and North Pacific (NP) from 2012 to 2019 are used for training and test. The best-track data are obtained from the International Best Track Archive for Climate Stewardship (IBTrACS). Simulation results show that comparing with the MFDL, the MVFF network improves the estimation accuracy, in terms of achieving TC intensity estimations with the average mean absolute errors (MAE) of from 4.37 m/s to 3.91 m/s for the NA TCs, and 4.92 m/s to 4.06 m/s for the NP TCs from 2018 to 2019, respectively.

Cloud detection is an important and challenging problem in atmospheric remote sensing. This study improves the cloud detection performance of the previously proposed cross-track infrared sounder (CrIS) full spectrum resolution (FSR) cloud detection index (FCDI) by changing the channel pairing criteria and using a new feature refining attention network (FRAnet). Instead of using the standard deviation of the simulated CrIS channel's brightness temperatures (BTs) as pairing index, this study adopts the root mean square error (RMSE) of the fitting results of the candidate pairs to select channel pairs. Since the channel's BT standard deviation mainly reflects the channel's reliability while the RMSE can directly reflect the pairing and fitting quality. Different RMSE thresholds are specified to represent the physical characteristics of different band groups, i.e., the long wave to short wave (LW-SW), LW to medium wave (MW), and MW-SW. After the new channel pairs being selected, the attention mechanism based FRAnet is applied to optimal utilize the FCDIs obtained from over 200 pairs. 40-day data is used for this study, and a precise algorithm is applied to match the CrIS data to the visible infrared imaging radiometer suite (VIIRS) data. As a result, new cloud labels are used. Simulation results show a largely improved cloud detection accuracy from the previously achieved ~ 80% to ~87.8%.

Chinese Fengyun 3 (FY-3) polar-orbiting satellites carry Microwave Humidity Sounders (MWHSs), including MWHS and MWHS-2 which are on board FY-3A/B and FY-3C/D, respectively. Understanding the quality of MWHS data is important for data assimilation and other applications. Examination of observed and simulated brightness temperatures and comparison with those of the NOAA-18 Microwave Humidity Sounder (MHS) reveal that the FY-3C MWHS-2 observations contain significant along-scanline noise. Similar noise exists in the humidity sounder data from FY-3 series satellites. In this study, the principal component analysis (PCA) method is used to identify and characterize the along-scanline noise, and a noise filter is also applied to the FY-3 series MWHS data by combining a PCA with a five-point smoother. The observation minus background or O-B biases of MWHS channels vary more smoothly with scan position after applying the filter, indicating that the along-scanline noise has been effectively reduced. Comparisons of O-B biases among all four FY-3 MWHS instruments show that biases for different channels have different asymmetric scan-angle features. As for along-scanline noise, FY-3A MWHS has the largest noise: 0.16, 0.27, and 0.89 K for channels 3–5, respectively. Noise in the FY-3D humidity channels is the smallest, of the order of 0.06–0.07 K. The along-scanline noise is also strongly correlated between channels.

Advanced Baseline Imager (ABI) aboard GOES-16 geostationary satellite represents the new generation of GOES imager instruments. It can provide more channels, higher spatial resolution and observation frequency than any previously launched GOES imagers. In this study, impacts of assimilating ABI radiance data at different model horizontal resolutions on quantitative precipitation forecast (QPF) skill over the continental U.S. are examined. The National Centers for Environmental Prediction (NCEP) Gridpoint Statistical interpolation (GSI) analysis system and Advanced Weather Research and Forecast model (WRF-ARW) are employed. After removing cloudy and precipitation-affected brightness temperatures through using an infrared (IR) cloud algorithm and implementing a bias correction, positive impacts are obtained for ABI data assimilation to improve QPFs associated with a typical spring precipitation case. Higher resolution experiments (6 km) improve the QPFs more significantly for heavier rainfall than coarser resolution experiments (15 km).

The launches of the US Advanced Baseline Imager (ABI) on November 19, 2016 and the Japanese Advanced Himawari Imager (AHI) on October 7, 2014 represented a new era of geostationary operational environmental satellite (GOES) imagers. Both ABI and AHI provide many more channels than any other already launched GOES imagers in the world. In this study, we compare the impacts of assimilating all ABI and/or AHI versus GOES-like infrared channels radiances on regional quantitative precipitation forecasts (QPFs) over US and China, respectively. Positive impacts on QPFs over land are obtained for a typical summer precipitation case by assimilating either all ten AHI infrared channels (AHI-10) or only four GOES-like AHI channels (AHI-4) when compared with a control experiment without assimilating GOES observations (CTRL). It is found that a southwest to northeast oriented band of the atmosphere with high water vapor content that was formed and moved inland with time under the influence of a subtropical high and an eastward propagating middle-latitude trough was responsible for the persistent precipitation in the eastern China of the selected case. The AHI-10 experiment generated larger improvements on QPFs due to it generating a wetter atmosphere in the middle and low troposphere over the ocean off the southeast coast of China than the AHI-4 and CTRL experiments.

Microwave Humidity Sounder (MWHS) onboard the Chinese FengYun-3C satellite has a total of 12 channels. Eight of these channels are located near the 118-GHz oxygen absorption band for probing atmospheric temperature and humidity fields from space. While the water vapor sounding channels near 183 GHz have been extensively studied in the past, we report the first satellite observations at 118 GHz. In this paper, the MWHS calibration accuracy is assessed by comparing the satellite observations with simulations. Using the collocated Constellation Observing System for Meteorology, Ionosphere, and Climate (COSMIC) radio occultation (RO) data in clear-sky conditions as inputs to Community Radiative Transfer Model, MWHS brightness temperatures are simulated for the five upper level sounding channels 2-9 located near the 118-GHz oxygen absorption band. For quality control of clear-sky radiance, a new cloud index is first developed based on the two MWHS window channels 1 and 10. Monthly mean biases of the antenna brightness temperature observations for the 118-GHz sounding channels are quantified by using more than 2000-5000 collocated COSMIC and MWHS data from August 2014 to February 2015. It is found that the bias of MWHS data relative to COSMIC RO simulation is dependent on channel and ranges within ±1.5 K. The standard deviation from the bias is the largest at MWHS channel 2, indicating a larger variability of the measurements of channel 2 than the other channels.

In the above-named article [ibid., vol. 50, no. 12, pp. 4892-4902, Dec. 2012], Fig. 6 is incorrect. The correct one is printed here.

MicroWave Humidity Sounder (MWHS) onboard both Fengyun-3A (FY-3A) and FY-3B satellites have three channels (channels 3-5) near the 183-GHz water-vapor absorption line. These channel frequencies are also used in other instruments such as Advanced Microwave Sounding Unit-B (AMSU-B) and Microwave Humidity Sounder (MHS) onboard MetOp and NOAA satellites. Both MWHS and MHS are cross-track scanners. In this paper, a comparison between the simulated brightness temperatures with MWHS measurements clearly shows that MWHS observations from the three sounding channels contain a scan-angle-dependent cohesive noise along the instrument scanline. This noise does not cancel out when a large amount of data over a sufficiently long period of time is averaged, which eliminates the possibility of such a noise to arise from the natural variability of the atmosphere and the surface. The noises are around 0.3, 0.2, and 0.2 K for channels 3-5, respectively. A principle component analysis is used for the characterization of this cohesive noise using one-month FY-3B MWHS data. It is shown that the MWHS cohesive noise is primarily contained in the first principal component (PC) mode, which mainly describes a scan-angle-dependent brightness temperature variation, i.e., a unique feature of the cross-tracking instrument. The first PC accounts for more than 99.91 % total variance in the three MWHS sounding channels. A five-point smoother is then applied to the first PC, which effectively removes such a data noise in the MWHS data. The reconstruction of the MWHS radiance spectra using the noise-filtered first PC component is of good quality. The scan-angle-dependent bias from the reconstructed MWHS data becomes more uniform and is consistent with the NOAA-18 MHS data.

The MicroWave Radiation Imager (MWRI) onboard the FengYun (FY)-3B satellite has five frequencies at 10.65, 18.7, 23.8, 36.5, and 89.0 GHz, each having dual channels at vertical and horizontal polarization states, respectively. It is found that radio-frequency interference (RFI) is present in MWRI data over land. The RFI signals are, in general, detectable from a spectral difference method and a principal component analysis (PCA) method. In particular, the PCA method is applied to derive RFI signals from natural radiations by using the characteristics of natural radiation measurements having all-channel correlations. In the area where data have a higher projection onto the first principle component (PC) mode, RFI is, in general, present. However, both the spectral and PCA methods cannot detect RFI reliably over frozen grounds and scattering surfaces, where the brightness temperature difference between 10.65 and 18.7 GHz is large. Thus, detection is improved through the use of normalized PCA. The new RFI detection algorithm is now working reliably for MWRI applications. It is found that RFI at 10.65 GHz distributes widely over Europe and Japan, and is less popular over the United States and China.