Land surface temperature (LST) plays a critical role in land surface processes. However, as one of the effective means for obtaining global LST observations, remote sensing observations are inherently affected by cloud cover, resulting in varying degrees of missing data in satellite-derived LST products. Here, we propose a solution. First, the data interpolating empirical orthogonal functions (DINEOF) method is used to reconstruct invalid LSTs in cloud-contaminated areas into ideal, clear-sky LSTs. Then, a cumulative distribution function (CDF) matching-based method is developed to correct the ideal, clear-sky LSTs to the real LSTs. Experimental results prove that this method can effectively reconstruct missing LST data and guarantee acceptable accuracy in most regions of the world, with RMSEs of 1–2 K and R values of 0.820–0.996 under ideal, clear-sky conditions and RMSEs of 4–7 K and R values of 0.811–0.933 under all weather conditions. Finally, a spatiotemporally continuous MODIS LST dataset at 0.05° latitude/longitude grids is produced based on the above method.

Convolutional neural networks (CNNs) are very important deep neural networks for analyzing visual imagery. However, most CNN-based methods have the problem of oversmoothing at boundaries, which is unfavorable for hyperspectral image classification. To address this problem, a spectral-spatial multiscale residual network (SSMRN) by fusing two separate deep spectral features and deep spatial features is proposed to significantly reduce over-smoothing and effectively learn the features of objects. In the implementation of the SSMRN, a multiscale residual convolutional neural network is proposed as a spatial feature extractor and a band grouping-based bi-directional gated recurrent unit is utilized as a spectral feature extractor. Considering that the importance of spectral and spatial features may vary depending on the spatial resolution of images, we combine both features with two weighting factors with different initial values that can be adaptively adjusted during the network training. To evaluate the effectiveness of the SSMRN, extensive experiments are conducted on public benchmark data sets. The proposed method can retain the detailed boundary of different objects and yield competitive results compared with several state-of-the-art methods.

Abstract. Runoff estimations play an important role in water resource planning and management. Existing hydrological models can be divided into physical models and data-driven models. Although the physical model contains certain physical knowledge and can be well generalized to new scenarios, the application of physical models is limited by the high professional knowledge requirements, difficulty in obtaining data and high computational costs. The data-driven model can fit the observed data well, but the estimation may not be physically consistent. In this letter, we propose a hybrid physical data (HPD) model combining physical model and deep learning model for runoff estimation. The model uses the output of a physical hydrological model together with the driving factors as another input of the neural network to estimate the monthly runoff of the upper Heihe River Basin in China. We show that the use of the HPD model improves the quality of runoff estimation, and results in high R2, NSE values of 0.969, and a low RMSE value of 9.645. It is indicated that the new model had an excellent learning capability to simulate runoff and flexible ability to extract complex relevant information; At the same time, the estimation capacity of peak runoff is optimized.

Runoff estimations play an important role in water resource planning and management. Many accomplishments have been made in runoff estimations based on data recorded at meteorological stations; however, the advantages of the use of remotely sensed data in estimating runoff in watersheds for which data are lacking remain to be investigated. In this study, the MOD13A2 normalized difference vegetation index (NDVI), TRMM3B43 precipitation (P), MOD11A2 land–surface temperature (LST), MOD16A2 evapotranspiration (ET) and hydrological station data were used as data sources with which to estimate the monthly runoff through the application of a fully connected long short–term memory (LSTM) model in the upstream reach of the Heihe River basin in China from 2001 to 2016. The results showed that inputting multiple remote sensing parameters improved the quality of runoff estimation compared to the use of rain gauge observations; an increase in R2 from 0.91 to 0.94 was observed from the implementation of this process, and Nash–Sutcliffe efficiency (NSE) showed an improvement from 0.89 to 0.93. The incorporation of rain gauge data as well as satellite data provided a slight improvement in estimating runoff with a respective R2 value of 0.95 and NSE value of 0.94. This indicates that the LSTM model based on remote sensing data has great potential for runoff estimation, and data obtained by remote sensing technology provide an alternative approach for estimating runoff in areas for which available data are lacking.

River runoff simulation and prediction are important for controlling the water volume and ensuring the optimal allocation of water resources in river basins. However, the instability of medium- and long-term runoff series increases the difficulty of runoff forecasting work. In order to improve the prediction accuracy, this research establishes a hybrid deep learning model framework based on variational mode decomposition (VMD), the mutual information method (MI), and a long short-term memory network (LSTM), namely, VMD-LSTM. First, the original runoff data are decomposed into a number of intrinsic mode functions (IMFs) using VMD. Then, for each IMF, a long short-term memory (LSTM) network is applied to establish the prediction model, and the MI method is used to determine the data input lag time. Finally, the prediction results of each subsequence are reconstructed to obtain the final forecast result. We explored the predictive performance of the model with regard to monthly runoff in the upper Heihe River Basin, China, and compared its performance with other single and hybrid models. The results show that the proposed model has obvious advantages in terms of the performance of point prediction and interval prediction compared to several comparative models. The Nash–Sutcliffe efficiency coefficient (NSE) of the prediction results reached 0.96, and the coverage of the interval prediction reached 0.967 and 0.908 at 95% and 90% confidence intervals, respectively. Therefore, the proposed model is feasible for simulating the monthly runoff of this watershed.