Three-dimensional cloud retrievals are critical for understanding their impact on climate and other applications, such as aviation safety, weather prediction, and remote sensing. However, obtaining high-resolution and accurate vertical representation of clouds remains unsolved due to the limitations imposed by satellite instrumentation, viewing conditions, and the complexity of cloud dynamics. Cloud masks are essential for comprehending various cloud vertical properties, but deriving accurate 3-D cloud masks from 2-D satellite imagery data is a challenging task. To tackle these challenges, we introduce a physics-informed loss function for training deep learning models that can extend 2-D cloud images into 3-D cloud masks. The proposed loss, called CloudMask loss, is composed of two domain knowledge-informed loss terms: one for evaluating cloud position and thickness and the other for measuring the number of layers. By combining these loss terms, we improve the trainability of the deep learning models for more accurate and meaningful results. We apply the proposed loss function to different neural networks and demonstrate significant improvements in multilayer cloud mask reconstruction. Utilizing the same neural network architecture, our proposed loss outperforms standard binary cross-entropy (BCE) loss in terms of multilayer cloud classification accuracy, number of layers accuracy, and thickness mean absolute error (MAE). The proposed loss function can be readily integrated into various neural network architectures, resulting in substantial performance gains in 3-D cloud mask generation.

Precipitation type is a key parameter used for better retrieval of precipitation characteristics as well as to understand the cloud–convection–precipitation coupling processes. Ice crystals and water droplets inherently exhibit different characteristics in different precipitation regimes (e.g., convection, stratiform), which reflect on satellite remote sensing measurements that help us distinguish them. The Global Precipitation Measurement (GPM) Core Observatory’s microwave imager (GMI) and dual-frequency precipitation radar (DPR) together provide ample information on global precipitation characteristics. As an active sensor, the DPR provides an accurate precipitation type assignment, while passive sensors such as the GMI are traditionally only used for empirical understanding of precipitation regimes. Using collocated precipitation type flags from the DPR as the “truth”, this paper employs machine learning (ML) models to train and test the predictability and accuracy of using passive GMI-only observations together with ancillary information from a reanalysis and GMI surface emissivity retrieval products. Out of six ML models, four simple ones (support vector machine, neural network, random forest, and gradient boosting) and the 1-D convolutional neural network (CNN) model are identified to produce 90–94% prediction accuracy globally for five types of precipitation (convective, stratiform, mixture, no precipitation, and other precipitation), which is much more robust than previous similar effort. One novelty of this work is to introduce data augmentation (subsampling and bootstrapping) to handle extremely unbalanced samples in each category. A careful evaluation of the impact matrices demonstrates that the polarization difference (PD), brightness temperature (Tc) and surface emissivity at high-frequency channels dominate the decision process, which is consistent with the physical understanding of polarized microwave radiative transfer over different surface types, as well as in snow and liquid clouds with different microphysical properties. Furthermore, the view-angle dependency artifact that the DPR’s precipitation flag bears with does not propagate into the conical-viewing GMI retrievals. This work provides a new and promising way for future physics-based ML retrieval algorithm development.

The correlation between MODIS aerosol optical thickness AOT and PM2.5 concentration in Fuling District of Chongqing was studied. The results showed that AOT and PM2.5 had significant seasonal variation characteristics, and the daily mean correlation Which was significant in the summer and poor correlation in winter. The influencing factors affecting the correlation between MODIS AOT and PM2.5 were studied, and the correlation coefficient between them was also greatly improved. Therefore, it is of great value to use MODIS AOT to monitor the distribution of PM2.5 in atmospheric particulate pollutants.

The Moderate Resolution Imaging Spectroradiometer (MODIS) Collection 6 (C6) aerosol optical depth (AOD) products from the 10/3 km Dark Target (DT) and Deep Blue (DB) algorithms are firstly evaluated using ground observed AODs by the sun photometer in Chongqing, a mountainous mega-city in southwest China. The validation results show that MODIS AODs from 10/3 km DT algorithm are comparable with those of the sun photometer, although there are slight overestimations. However, the DB algorithm substantially underestimates MODIS AODs when comparing with those of the sun photometer. Error analyses imply that the bias of surface reflectance estimation is the main error source for both algorithms. The cloud screening scheme of the DT algorithm is more effective than the DB algorithm. The cloud vicinity effect should be considered in the quality control processes for both of the algorithms. A sensitivity test suggests that in complex terrain area, like Chongqing, the collocation method in the validation of satellite products should be carefully selected according to local circumstances. When comparing the monthly mean AODs of MODIS products with sun photometer observations, it shows that the Terra MODIS AOD products are valid to represent the mean statuses in summer and autumn, but the monthly mean of Aqua MODIS AODs are limited in Chongqing.

The pine wilt disease is a devastating disease of pine trees. In China, the first discoveries of the pine wilt disease on 1982 at Dr. Sun Yat-sen's Mausoleum in Nanjing. It occurred an area of 77000 hm2 in 2005, More than 1540000 pine trees deaths in the year. Many districts of Chongqing in Three Gorges Reservoir have different degrees of pine wilt disease occurrence. It is a serious threat to the ecological environment of the reservoir area. Use unmanned airship to carry high spectrum remote sensing monitoring technology to develop the study on pine wood nematode disease early diagnosis and early warning and forecasting in this study. The hyper spectral data and the digital orthophoto map data of Fuling District Yongsheng Forestry had been achieved In September 2015. Using digital image processing technology to deal with the digital orthophoto map, the number of disease tree and its distribution is automatic identified. Hyper spectral remote sensing data is processed by the spectrum comparison algorithm, and the number and distribution of disease pine trees are also obtained. Two results are compared, the distribution area of disease pine trees are basically the same, indicating that using low air remote sensing technology to monitor the pine wood nematode distribution is successful. From the results we can see that the hyper spectral data analysis results more accurate and less affected by environmental factors than digital orthophoto map analysis results, and more environment variable can be extracted, so the hyper spectral data study is future development direction.

The TanSat carbon satellite is to be launched at the end of 2016. In order to verify the performance of its instruments, a flight test of TanSat instruments was conducted in Jilin Province in September, 2015. The flight test area covered a total area of about 11,000 km2 and the underlying surface cover included several lakes, forest land, grassland, wetland, farmland, a thermal power plant and numerous cities and villages. We modeled the column-average dry-air mole fraction of atmospheric carbon dioxide (XCO2) surface based on flight test data which measured the near- and short-wave infrared (NIR) reflected solar radiation in the absorption bands at around 760 and 1610 nm. However, it is difficult to directly analyze the spatial distribution of XCO2 in the flight area using the limited flight test data and the approximate surface of XCO2, which was obtained by regression modeling, which is not very accurate either. We therefore used the high accuracy surface modeling (HASM) platform to fill the gaps where there is no information on XCO2 in the flight test area, which takes the approximate surface of XCO2 as its driving field and the XCO2 observations retrieved from the flight test as its optimum control constraints. High accuracy surfaces of XCO2 were constructed with HASM based on the flight’s observations. The results showed that the mean XCO2 in the flight test area is about 400 ppm and that XCO2 over urban areas is much higher than in other places. Compared with OCO-2’s XCO2, the mean difference is 0.7 ppm and the standard deviation is 0.95 ppm. Therefore, the modelling of the XCO2 surface based on the flight test of the TanSat instruments fell within an expected and acceptable range.

CO2 is the greenhouse gas that influenced by human activities most and has a great impact on the climate change. Monitoring of global CO2 variations on a basis of, high precision has great significance for the study of global climate change and carbon cycle, as well as the understanding of CO2 sources and sinks. This study developes the forward model and inversion software system GF_VRTM-V1.0 for space-borne near-infrared hyperspectral measurements of CO2 into a ground-based observation version. The simulation results are compared with the ground-based data of the column-averaged mole fraction of Carbon Dioxide measured at Fuling, and showed general agreement.

The spectra of O2 A-band (0.76 μm) and CO2 near-infrared emissions (1.6 μm) for Medium-resolution Satellite (SCIAMACHY) are simulated by the SCIATRAN model (V3.1.29), and compared with the ESFT and LBL method, as the inversion accuracy and time consuming. The time consuming of LBL was more than ESFT with the relative error less than 1%, especially for the CO2 band. But for the CO2 (2.0 um) of High-resolution Satellite, the opposite result was found. That is to say, the LBL method was more suitable for High-resolution Satellite. Different wavelength intervals and integral wavelength steps are applied to the LBL to select the most appropriate combination for High-resolution SatelliteO2 A-band (0.76 μm) and CO2 near-infrared band (1.58 μm).

The greenhouse gas carbon dioxide, for which short-wave infrared remote sensing detection is carried out by using satellite sensors to measure the Earth’s atmosphere scattering solar radiation, and makes use of the inversion algorithm to achieve measurements. Most of the solar radiation enter the satellite sensors after surface reflection, so the surface albedo which reflects the surface features is one of the important parameters which affect the accuracy of the detection. Aiming at the great demands of high precision carbon dioxide for greenhouse gas, this study first investigate the effects of the Earth’s surface albedo on the observed spectra. Simulation results show that the increase in the surface albedo will enhance the observed spectral intensity, especially larger in the O2-A band than in the 1.6 μm band. In other words, the surface albedo has a greater impact on O2-A ban. In the actual satellite inversio, the surface types of actual observation pointare uncertain, which will result in the error of surface albedo. Effect of surface albedo on the inverted XCO2 is analyzed when the surface albedo is changed by changing the type of surfac. Two observation cases are analyzed in detail. One is on April 23, 2009 for the desert surface, and another on May 21, 2013 for the grass surfac. Results show that when the O2-A band surface albedo approximates to the real surface albedo valu, the relative error of the inverted XCO2 is the smaller. If the relative changes of the O2-A band surface albedo exceed 0.25 in the grass surfac or 0.35 in the desert surface, the relative error of the inverted XCO2 will be greater than 1%, not satisfying the design requirement of the inversion system. In contrast, the changesin 1.6 μm band surface albedo have negligible effect on the inverted XCO2. This study shows the importance of surface albedo in the process of satellite remote sensin, and provides an important theoretical basis and guidance for improving the accuracy of remote sensing detectio. All these are significantly contributed to the hyperspectral satellite observation of the greenhouse gas, the investigation of global CO2 distributions, and the prediction and monitoring of the climate change.

An innovative instrument based on Savart polariscope is proposed to simultaneously acquire image, spectrum, and full Stokes parameters of a target. The principle, scheme, and technique are discussed, and some preliminary results are presented.

The atmospheric carbon dioxide CO2 column concentrations observed by the Greenhouse Gases Observing Satellite GOSAT and ground stations at Mt Waliguan 36.29° N, 100.90° E and Lulin 23.47° N, 120.87° E in China are compared. The data covered time periods from June 2009 to November 2011 for GOSAT and from July 2009 to December 2010 for the ground stations. The GOSAT monthly mean data tend to be generally smaller than those of the ground measurements by 5–10 ppm. The spatial and temporal variations of the atmospheric XCO2 dry air, column averaged, molar fraction of CO2 concentrations, especially in the regions of China, are analysed by using the GOSAT monthly mean data. The variations are more significant in the northern hemisphere than in the southern hemisphere and show relatively high values and obvious fluctuations in the 15° N–45° N latitudinal band. These are generally consistent with the measurements of the Scanning Imaging Absorption Spectrometer for Atmospheric Cartography SCIAMACHY and Atmospheric Infrared Sounder AIRS. The satellite data show significant seasonal variations, with maximum in April and May and minimum in September and October. This feature is in general agreement with that of the ground observations and previous reports. In the regions of China, the XCO2 ranged from 355 ppm to 385 ppm with a mean of 374 ppm, which is in agreement with the global concentration.

the history, status and the future of the passive measurement technology used for upper atmosphere measurement (uam) are introduced. the principle and method for uam are introduced. the emission source, four intensities method, and limb measurement model for uam are presented. the design and manufacture of field-widened, achromatic, temperature compensation wind imaging interferometer (fatwindii) developed are described emphatically. the laboratory experimental system for simulation measurement of fatwindii is built up. based on this experimental system, some simulation measurement results of velocity, temperature and pressure of upper atmosphere are presented in this paper. the research has scientific significance and practical value for physical geography, atmospheric science, environment protection, national defense and national economic construction.

The instrument forward model of the modified super-wide-Sagnac imaging interferometer based on liquid crystals on Silicon (MSASII-LCoS) is developed as an integrated code package with Matlab language to simulate the images of satellite observations. There are five sub-models in the forward model including radiation model of O(1S), orbit attitude, filter, interferometer and array detector. The principle of each sub-model is described separately and then the overall forward model equation is derived. The four simulation images are obtained. Based on the integrated signal level for the daytime observations, the apparent measurement error of wind is less than 3m/s and the signal noise ratio (SNR) is greater than 194 with a binning of 2×25pixels at the tangent height range of 70–190km.

The spatio–temporal variations of major meridional modes are studied by using the ECMWF and NCEP/NCAR reanalyzed geopotential data between 70 and 10hPa during 1979 and 2001. The variance contribution rates from the first and second modes are 56–69% and 14–22%, respectively, for ECMWF, but 76–85% and 9–10% for NCEP/NCAR. The climatic trend coefficients are positive and negative in the troposphere and stratosphere, respectively. The reversal is remarkably correlated with the AO/NAM, AAO/SAM and polar vortex, suggesting their important role in connecting the mid-low and upper circulation and the interaction between SH and NH.

The Michelson Interferometer for Passive Atmospheric Sounding (MIPAS) is operating on the ESA ENVIronmental SATellite since March 2002. The high resolution (0.025 cm -1 ) limb-emission measurements acquired by MIPAS in the first two years of operations constitute a self-consistent set of data with very good geographical and time coverage. These measurements have been re-processed by ESA up to Level 2, with the most recent versions of both Level 1b and Level 2 algorithms. The products of the ESA Level 2 algorithm are geolocated profiles of temperature and of volume mixing ratio of six key atmospheric constituents: H2O, O3, HNO3, CH4, N2O and NO2. As for all the measurements made with innovative instruments and techniques, this data set requires a thorough validation. During the last year, a large team of scientists spent great efforts in the validation of this data set. In particular, the authors of this paper have focused their activities on the validation of temperature. The validation was carried-out by comparing MIPAS retrieved temperature with correlative measurements made by radiosondes, lidars, in-situ and remote sensors operated either from stratospheric balloons or satellites. Preliminary results show that MIPAS profiles are affected by a bias generally consistent with their a-priori estimated systematic error component. This bias is usually localized at the edges of the altitude range covered by the MIPAS scan (6 - 68 km).

This analysis presents comparisons of the atmospheric temperatures retrieved from CPS/SAC-C radio occultation observations using the JPL retrieval software, and from MIPAS/ENVISAT infrared spectrum measurements using the IMK data processor. Both individual profiles and zonal means of the atmospheric temperature at different seasons and geo-locations show reasonable agreement. For the temperatures at altitudes between 8-30 km, the mean differences between the correlative measurements are estimated at less than 2 K with rms deviations less than 5 K. A similar cross comparison technique can be used to help validate the observed temperatures from the new EOS MLS instrument, to be launched in 2004.

The temperatures retrieved from MIPAS/ENVISAT limb mid-infrared emission and CHAMP GPS radio occultation measurements are compared at. altitudes between 8 - 30 km during the stratospheric major sudden warming in the southern hemisphere winter of 2002. The mean differences between the correlative measurements of the two instruments are less than similar to1 K with rms deviationS of similar to3-5 K. The MIPAS temperatures are slightly higher than those of GPS-RO around 30 km. Possible explanation is discussed.

Two radio Occultation (RO) retrieval schemes designed at IGAM to enhance the performance at high altitudes are presented, applied to CHAMP RO data. and validated against ECMWF analyses, GFZ operational retrieval, ENVSAT/MIPAS-, and ENVIAT/GOMOS-derived temperature profiles. IGAM proposes to include background information into the RO retrieval only at one point at bending angle level in order to be able to track error characteristics of the retrieved product. The results show very good agreement with GFZ retrieval and ECMWF analysis below 15 km and. depending on the background information used, either a significant warm bias or essentially no bias up to 30 km. Compared to MIPAS, the only independent data source. the IGAM/ECMWF retrieval is unbiased up to 40 km.