Drought is a complex hazard, and it has an impact on agricultural, ecological, and socio-economic systems. The vegetation condition index (VCI), which is derived from remote-sensing data, has been widely used for drought monitoring. However, VCI based on the normalized difference vegetation index (NDVI) does not perform well in certain circumstances. In this study, we examined the utility of the vegetation index based on the universal pattern decomposition method (VIUPD) based VCI for drought monitoring in various climate divisions across the continental United States (CONUS). We compared the VIUPD-derived VCI with the NDVI-derived VCI in various climate divisions and during different sub-periods of the growing season. It was also compared with other remote-sensing-based drought indices, such as the temperature condition index (TCI), precipitation condition index (PCI) and the soil moisture condition index (SMCI). The VIUPD-derived VCI had stronger correlations with long-term in situ drought indices, such as the Palmer Drought Severity Index (PDSI) and the standardized precipitation index (SPI-3, SPI-6, SPI-9, and SPI-12) than did the NDVI-derived VCI, and other indices, such as TCI, PCI and SMCI. The VIUPD has considerable potential for drought monitoring. As VIUPD can make use of the information from all the observation bands, the VIUPD-derived VCI can be regarded as an enhanced VCI.

We examined the relationship between satellite measurements of solar-induced chlorophyll fluorescence (SIF) and several meteorological drought indices, including the multi-time-scale standard precipitation index (SPI) and the Palmer drought severity index (PDSI), to evaluate the potential of using SIF to monitor and assess drought. We found significant positive relationships between SIF and drought indices during the growing season (from June to September). SIF was found to be more sensitive to short-term SPIs (one or two months) and less sensitive to long-term SPI (three months) than were the normalized difference vegetation index (NDVI) or the normalized difference water index (NDWI). Significant correlations were found between SIF and PDSI during the growing season for the Great Plains. We found good consistency between SIF and flux-estimated gross primary production (GPP) for the years studied, and synchronous declines of SIF and GPP in an extreme drought year (2012). We used SIF to monitor and assess the drought that occurred in the Great Plains during the summer of 2012, and found that although a meteorological drought was experienced throughout the Great Plains from June to September, the western area experienced more agricultural drought than the eastern area. Meanwhile, SIF declined more significantly than NDVI during the peak growing season. Yet for senescence, during which time the reduction of NDVI still went on, the reduction of SIF was eased. Our work provides an alternative to traditional reflectance-based vegetation or drought indices for monitoring and assessing agricultural drought.

Landsat 8, the most recently launched satellite of the series, promises to maintain the continuity of Landsat 7. However, in addition to subtle differences in sensor characteristics and vegetation index (VI) generation algorithms, VIs respond differently to the seasonality of the various types of vegetation cover. The purpose of this study was to elucidate the effects of these variations on VIs between Operational Land Imager (OLI) and Enhanced Thematic Mapper Plus (ETM+). Ground spectral data for vegetation were used to simulate the Landsat at-senor broadband reflectance, with consideration of sensor band-pass differences. Three band-geometric VIs (Normalized Difference Vegetation Index (NDVI), Soil-Adjusted Vegetation Index (SAVI), Enhanced Vegetation Index (EVI)) and two band-transformation VIs (Vegetation Index based on the Universal Pattern Decomposition method (VIUPD), Tasseled Cap Transformation Greenness (TCG)) were tested to evaluate the performance of various VI generation algorithms in relation to multi-sensor continuity. Six vegetation types were included to evaluate the continuity in different vegetation types. Four pairs of data during four seasons were selected to evaluate continuity with respect to seasonal variation. The simulated data showed that OLI largely inherits the band-pass characteristics of ETM+. Overall, the continuity of band-transformation derived VIs was higher than band-geometry derived VIs. VI continuity was higher in the three forest types and the shrubs in the relatively rapid growth periods of summer and autumn, but lower for the other two non-forest types (grassland and crops) during the same periods.

As a kind of soil-borne epidemic disease, bacterial wilt (BW) is one of the most serious diseases in tomatoes in southern China, which may significantly reduce food quality and the total amount of yield. Hyperspectral remote sensing can detect crop diseases in the early stages and offers potential for BW detection in tomatoes. Tomatoes in southern China are commonly cultivated in greenhouses or bird nets, limiting the application of remote sensing based on natural sunlight. To resolve these issues, we collected the spectrum of tomatoes firstly using the HS-VN1000B Portable Intelligent Spectrometer, which is equipped with a simulated solar light source. We then proposed a tomato BW detection model based on some optimal spectral features. Specifically, these optimal features, including vegetation indexes and principal components (PCs), were extracted by the sequential forward selection (SFS), the simulated annealing (SA), and the genetic algorithm (GA) and were finally fed into the support vector machine (SVM) classifier to detect diseased tomatoes. The results showed that the infected and healthy tomatoes exhibit different spectral characteristics for both leave and stem spectra, especially for near-infrared bands. In addition, the BW detecting model built by the combination of GA and SVM (GA-SVM) achieved the best performance with overall accuracies (OA) of 90.7% for leaves and 92.6% for stems. Compared with the results based on leaves, spectral features of stems provided better accuracy, indicating that the symptom of early infection of BW is more significant in tomato stems than in leaves. Further, the reliability of the GA-SVM tomato stem model was verified in our 2022 experiment with an OA of 88.6% and an F1 score of 0.80. Our study provides an effective means to detect BW disease of tomatoes in the early stages, which could help farmers manage their tomato production and effectively prevent pesticide abuse.

Deep learning methods used for hyperspectral image (HSI) classification often achieve greater accuracy than traditional algorithms but require large numbers of training epochs. To simplify model structures and reduce their training epochs, an end-to-end deep learning framework incorporating a spectral-spatial cascaded 3D convolutional neural network (CNN) with a convolutional long short-term memory (CLSTM) network, called SSCC, is proposed herein for HSI classification. The SSCC framework employs cascaded 3D CNN to learn the spectral-spatial features of HSIs and uses the CLSTM network to extract sequence features. Residual connections are used in SSCC to accelerate model convergence, with the outputs of previous convolutional layers concatenated as inputs for subsequent layers. Moreover, the data augmentation, parametric rectified linear unit, dynamic learning rate, batch normalization, and regularization (including dropout and L2) methods are used to increase classification accuracy and prevent overfitting. These attributes allow the SSCC framework to achieve good performance for HSI classification within 20 epochs. Three well-known datasets including Indiana Pines, University of Pavia, and Pavia Center were employed to evaluate the classification performance of the proposed algorithm. The GF-5 dataset of Anxin County, obtained from China’s recently launched spaceborne Advanced Hyperspectral Imager, was also used for classification experiments. The experimental results demonstrate that the proposed SSCC framework achieves state-of-the-art performance with better training efficiency than other deep learning methods.