Accurate radiometric calibration is crucial for satellite sensors, ensuring reliable remote sensing data. However, current calibration methods often rely on a limited number of samples, leading to uncertainties in calibration process. In this study, we presented a comprehensive approach for calibrating FY-3D MERSI-II sensor by leveraging an extensive dataset of over 110 000 dark pixel samples collected from three diverse oceanic regions. Through a meticulous data screening process, we identified suitable calibration samples to enhance the robustness of our approach. Employing a radiative transfer model, we conducted forward modeling to calibrate the sensor's observations. Notably, our coefficients of determination ($\mathit{{R}}^{2}$) establish the efficacy of the Rayleigh scattering method for channels CH09, CH10, and CH11, demonstrating a strong correlation between the spectral range of 443 and 555 nm. The calibration results revealed improved accuracy, with mean absolute BIAS values ranging from 1.329% to 11.265% across different seas, highlighting the wavelength-dependent uncertainty. Our approach effectively addresses the limitations associated with a limited number of referenced points, thereby enhancing the reliability of calibrated results. This research significantly contributes to advancing radiometric calibration for the FY-3D MERSI II sensor by explicitly emphasizing usage of over 110 000 dark pixels during the calibration process. These findings not only enhance the accuracy of satellite-based oceanic observations but also provide valuable insights into calibration methodologies for future studies.

The accuracy of fractional vegetation cover (FVC) estimation has important significance in high-precision agricultural and ecological environment assessments. However, the presence of shadows under natural light conditions can cause problems such as poor vegetation distinguishability and blurred boundaries during vegetation segmentation. This leads to missing or over-segmented vegetation in shaded areas, thus reducing the accuracy of FVC. Polarization information can reflect the texture structure features, edge characteristics and surface state information, providing important supplementary information to light intensity information. To address the issue of shadows interfering with vegetation segmentation, a Siamese coupling swin transformer (SiamC Transformer) is proposed in this study. In this network, a dual-flow structure is used to simultaneously retrieve the feature information of vegetation degree of linear polarization (DoLP) images and light intensity images simultaneously to obtain multi-dimensional global semantic information of multi-scale vegetation. In the feature fusion stage, two fusion modules are proposed: an adaptive fusion module (AFM) and an adaptive fusion plus module (AFPM), which can fuse shallow spatial information, texture information and deep semantic information. The AFM and AFPM enable the network to better achieve object localization, region activation and edge sharpening, while improving the segmentation accuracy of small objects. The experimental results show that the mean intersection over union (mIoU) of this network is as high as 97.46% on vegetation datasets consisting of light intensity and polarization images, which is better than that of other algorithms. This network has higher accuracy and adaptability for shadow scenes and improves the accuracy of FVC calculation under shadow conditions.

As the largest contiguous karst area of China, the southwestern karst area is a typical ecologically fragile area affecting local vegetation dynamics. Ecosystem water use efficiency (WUE) is an important factor outlining the vegetation’s ability to produce organic matter with a limited water supply. Therefore, it is important to determine WUE variation trends in this ecologically fragile region. In this paper, we used MODIS remote sensing datasets, meteorological data and land cover data to analyze the spatiotemporal changes in vegetation water use efficiency in the southwest karst region from 2001 to 2017. We also further quantitatively analyzed the impact of climate change and human activities on the spatial and temporal patterns of vegetation WUE in the study area. The main conclusions were as follows. (1) From 2001 to 2017, in terms of temporal characteristics, the interannual variation in WUE fluctuated greatly, ranging from 1.33 to 1.51 g C kg -1 H 2 O, with a multiyear average of 1.43 g C kg -1 H 2 O and an average rate of change of -0.0046 g C kg -1 H 2 O yr -1 . In terms of spatial characteristics, areas with a higher WUE were concentrated in central Sichuan and northeastern Yunnan. (2) There were also certain differences in the WUE for different vegetation types. The annual average WUE of each vegetation type decreased in the following order: evergreen coniferous forest> evergreen broad-leaved forest> mixed forest> deciduous broad-leaved forest> cultivated land> deciduous coniferous forest> grassland> cultivated land and natural vegetation> shrub forest. (3) The vegetation WUE of 70.66% in this area was positively correlated with temperature. Additionally, 79.68% of the vegetation WUE was negatively correlated with precipitation. The relative contribution rates of climate change and human activities to the change trend in WUE were 15% and 85%, respectively. These results provide scientific support for local vegetation restoration and protection policies.

As the largest contiguous karst area in China, the southwestern karst area is a typical ecologically fragile area affecting local vegetation dynamics. Ecosystem water use efficiency (WUE) is an important factor reflecting the ability of vegetation to produce organic matter with a limited water supply. Therefore, determining the WUE variation trends in this ecologically fragile region is important. In this paper, we used MODIS remote sensing datasets, meteorological data, and land cover data to analyze the spatiotemporal changes in vegetation water use efficiency in the southwestern karst region from 2001 to 2017. We also further quantitatively analyzed the effects of climate change and human activities on the spatial and temporal patterns of vegetation WUE in the study area. The main conclusions were as follows. (1) From 2001 to 2017, in terms of temporal characteristics, the interannual variation in WUE fluctuated greatly, ranging from 1.33 to 1.51 g C kg−1 H2O, with a multiyear average of 1.43 g C kg−1 H2O and an average rate of change of − 0.0046 g C kg−1 H2O year−1. In terms of spatial characteristics, areas with a higher WUE were concentrated in central Sichuan and northeastern Yunnan. (2) The annual average WUE of each vegetation type decreased in the following order: evergreen coniferous forest > evergreen broad-leaved forest > mixed forest > deciduous broad-leaved forest > cultivated land > deciduous coniferous forest > grassland > cultivated land and natural vegetation > shrub forest. (3) The vegetation WUE of 70.66% in this area was positively correlated with temperature. Additionally, 79.68% of the vegetation WUE was negatively correlated with precipitation. The relative contribution rates of climate change and human activities to the change trend in WUE were 15% and 85%, respectively. Compared with WUE results in other studies, the WUE of different karst landform areas obtained in this study was quite different, indicating that the geological and landform features of the karst area are complex. Our study provides scientific support for local vegetation restoration and protection policies and promotes the understanding of the principle of the carbon–water cycle in karst areas.

The increased growth of vegetation has the potential to slow global climate warming. Therefore, analyzing and predicting the response assessment of Chinese vegetation to climate change is of great significance to studies of global warming. In this paper, we examine the spatiotemporal dynamics of vegetation leaf area index (LAI) values in China from 1981 to 2017 and their correlations with meteorological (hydrothermal) factors based on trend analysis and correlation analysis. We further construct an LAI prediction model based on hydrothermal conditions. The climate data obtained under different scenarios in the CMIP5 and CMIP6 climate models were used to predict the dynamic change trend of vegetation LAI from 2021 to 2100. The results show that most areas of China (72.82%) showed an improving trend in vegetation LAI from 1981 to 2017, during which the annual average LAI value increased at a rate of 0.0029 year−1. Vegetation LAI in China was significantly correlated with climatic factors (temperature, precipitation, and evapotranspiration), and the LAI prediction model constructed based on hydrothermal conditions had a high accuracy (Pearson’s Cor value is 0.9729). From 2021 to 2100, approximately 2/3 of China’s vegetation LAI area showed an improvement trend, and the impact of climate change on vegetation LAI predictions under the high emission scenario was greater than that under the low emission scenario. This research can provide a basis for studies on the climatic drivers of vegetation change and the global vegetation dynamic model.