Accurate and timely estimation of crop yield at a small scale is of great significance to food security and harvest management. Recent studies have proven remote sensing is an efficient method for yield estimation and machine learning, especially deep learning, can infer a good prediction by integrating multi-source datasets like satellite data, climate data, soil data, and so on. However, there are some bottleneck challenges to improve accuracy. First, the popular remote sensing data used for yield prediction fall into two major groups, time-series data, and constant data. Surprisingly little attention has been devoted to deep learning networks that can integrate the two kinds of data effectively; second, both temporal and spatial features play a role in affecting the yields. But most of the existing approaches employed either CNN (convolutional neural network) or RNN (recurrent neural network). CNN cannot learn temporal patterns, while RNN barely can learn spatial characteristics. This work proposed a novel multi-level deep learning (MLDL) model coupling RNN and CNN to extract both spatial and temporal features. The inputs include both time-series remote sensing data and soil property data and the model outputs yield. We experimented with the model in U.S. Corn Belt states and used it to predict corn yield from 2013 to 2016 at the county-level. The results approve the effectiveness and advantages of the proposed approach over the other methods. In the future, the model will be used on other crops like soybean and winter wheat to assist agriculture decision making.

Accurate bathymetric data is crucial for marine and coastal ecosystems. A lot of studies have been carried out for nearshore bathymetry using satellite data. The approach adopted extensively in shallow water depths estimation has recently been one of empirical models. However, the linear empirical model is simple and only takes limited band information at each bathymetric point into consideration. It may be not suitable for complex environments. In this paper, a deep learning framework was proposed for nearshore bathymetry (DL-NB) from ICESat-2 LiDAR and Sentinel-2 Imagery datasets. The bathymetric points from the spaceborne ICESat-2 LiDAR were extracted instead of in situ measurements. By virtue of the two-dimensional convolutional neural network (2D CNN), DL-NB can make full use of the initial multi-spectral information of Sentinel-2 at each bathymetric point and its adjacent areas during the training. Based on the trained model, the bathymetric maps of several study areas were produced including the Appalachian Bay (AB), Virgin Islands (VI), and Cat Island (CI) of the United States. The performance of DL-NB was evaluated by empirical method, machine learning method and multilayer perceptron (MLP). The results indicate that the accuracy of the DL-NB is better than comparative methods can in nearshore bathymetry. After quantitative analysis, the RMSE of DL-NB could achieve 1.01 m, 1.80 m and 0.28 m in AB, VI and CI respectively. Given the same data conditions, the proposed method can be applied for high precise global scale and multitemporal nearshore bathymetric maps generation, which are beneficial to marine environmental change assessment and conservation.

Accurate estimation of GDP at small geographies is of great significance to evaluate the distribution and dynamics of socio-economic development. Nighttime light(NTL) data is becoming increasingly important in socio-economic data estimation. However, previous research has found that using NTL alone is insufficient to accurately measure the GDP at small geographies and the contribution of NTL for time-series GDP estimation is unreliable. This paper proposed a Deep Learning method for the CONUS time-series (2012-2015) GDP estimation at the county-level. The model is developed by combining the NTL data from the Visible Infrared Imaging Radiometer Suite Day/Night Band and the MODIS land cover data. The proposed method can improve the existing methods mainly in two ways. First, by taking advantage of the great computing power of the GEE, a histogram-based feature regulation method was employed, which not only keeps more information over regions but also provides dimension-reduced tensors from mass remote sensing data. Second, a Multi-Inputs Convolutional Neural Network-based model was proposed instead of the traditional linear regression model for multi-source feature exploration and learning. The proposed method was evaluated by leave-one-year-out cross-validation with the time-series(2012-2015) data. The results show that the $R^{2}$ between the actual and estimated GDP are 0.81, 0.83, 0.83, and 0.83 for years from 2012 to 2015, indicating a good predictive power of the proposed model. Given that the data employed are globally and publicly available, the proposed method would also be applicable in other countries or regions where socio-economic survey data is hard to obtain.

During the construction of deep foundation pits in subways, it is crucial to closely monitor the horizontal displacement of the pit enclosure to ensure stability and safety, and to reduce the risk of structural damage caused by pit deformations. With advancements in machine-learning (ML) techniques and correlation analysis in engineering, data-driven methods that combine ML with engineering monitoring data have become increasingly popular. These methods offer benefits such as high prediction accuracy, efficiency, and cost effectiveness. The main goal of this study was to develop a machine-learning method for predicting the enclosure deformation of deep foundation pits. This was achieved by analyzing the factors influencing deep foundation-pit enclosure deformation and incorporating historical cases and monitoring reports. The performance of each machine-learning prediction model was systematically analyzed and evaluated using K-Fold cross validation. The results revealed that the random forest model outperformed the other models. The result of the test data showed that the random forest model achieved an R² of 0.9905, an MAE of 0.8572 mm, and an RMSE of 1.9119 mm. Feature importance analysis identified the depth of enclosure structure, water level, surface settlement, axial force, and exposure time as the most critical factors for accurate prediction. The depth of the enclosure structure had an especially significant impact on the prediction of enclosure deformation.

The coral reef geomorphic zone classification (CRGZC) map can provide a wealth of information for coastal management and protection. Remote sensing plays an important role in CRGZC by virtue of its speed, wide range, and low cost. Although many excellent results have been achieved in this field, there are still some shortcomings. With the development of machine learning, such methods are gradually introduced to CRGZC, yet the research and application of deep learning methods are still relatively few. In this article, based on ICESat-2 data and multispectral images, a deep learning model coupled with convolutional neural network (CNN) and random forest (RF) was proposed for coral reef geomorphic zone classification (CR_CRGZC). First, the priori bathymetry points were extracted from ICESat-2. Then, a near-shore bathymetry map was generated using a log-ratio model. Finally, topographic data and multispectral images were combined to achieve CRGZC through CR_CRGZC. The northeastern part of Coffin Island (CI) and the southern part of Punta Vaquero (PV) in Puerto Rico Island were selected as study areas. By comparing the classification results with those of CNN, RF, and maximum likelihood classification, CR_CRGZC outperformed the other classification methods. By quantitative analysis, the OA and Kappa coefficients of CR_CRGZC were 91.91% and 0.9013 in the CI region; and 89.91% and 0.8735 in the PV region, respectively. Under the same environmental requirements, this approach can map high-precision submeter CRGZC maps, providing a database for dynamic coral reef habitat mapping, which contributes to marine coastal ecosystem protection and coastal underwater topography monitoring.

The selection of reasonable locations for urban wastewater treatment plants (WWTPs) is significant in controlling water pollution. However, most current WWTP location selection models focus only on population density, industrial zone scale and geographic elements, while ignoring water pollution and local socio-cultural indicators. We propose a novel indicator system from RS/GIS data to select reasonable locations for WWTPs in Phnom Penh to avoid water environmental deterioration and harm to human health. The frequency of occurrence of water eutrophication is derived from time series RS data and reveals the degree of pollution of local water ecosystems, and is included as a demand indicator for the first time. In order to respect local socio-cultural customs, buffer zones for religious sites are included in the indicator system for the first time to fully determine the zones where construction of WWTPs is forbidden. Subsequently, WWTP locations are selected with the fusion of the minimized facilities number and maximum coverage models. The results demonstrate that the selected locations are all highly suitable and outside zones where construction is forbidden. The method proposed in the article provides a more comprehensive and scientific perspective for WWTP location selection.