Next-level vegetation health index forecasting: A ConvLSTM study using MODIS Time Series.

The Vegetation Health Index (VHI) is a metric used to assess the health and condition of vegetation, based on satellite-derived data. It offers a comprehensive indicator of stress or vigor, commonly used in agriculture, ecology, and environmental monitoring for forecasting changes in vegetation health. Despite its advantages, there are few studies on forecasting VHI as a future projection, particularly using up-to-date and effective machine learning methods. Hence, the primary objective of this study is to forecast VHI values by utilizing remotely sensed images. To achieve this objective, the study proposes employing a combined Convolutional Neural Network (CNN) and a specific type of Recurrent Neural Network (RNN) called Long Short-Term Memory (LSTM), known as ConvLSTM. The VHI time series images are calculated based on the Normalized Difference Vegetation Index (NDVI) and Land Surface Temperature (LST) data obtained from the Moderate Resolution Imaging Spectroradiometer (MODIS) aboard the Terra and Aqua satellites. In addition to the traditional image-based calculation, the study suggests using global minimum and global maximum values (global scale) of NDVI and LST time series for calculating the VHI. The results of the study showed that the ConvLSTM with a 1-layer structure generally provided better forecasts than 2-layer and 3-layer structures. The average Root Mean Square Error (RMSE) values for the 1-step, 2-step, and 3-step ahead VHI forecasts were 0.025, 0.026, and 0.026, respectively, with each step representing an 8-day forecast horizon. Moreover, the proposed global scale model using the applied ConvLSTM structures outperformed the traditional VHI calculation method.

Monitoring land use and land cover change near a nuclear power plant construction site: Akkuyu case, Turkey.

Land use and land cover (LULC) change analysis of the construction site and its surroundings of the Akkuyu Nuclear Power Plant project in southern Turkey was undertaken in this case study, which was supported by remotely sensed Landsat 8 image composites. The composite images compiled in 2017 and 2021 were prepared on the Google Earth Engine platform. The Random Forest algorithm was used as the classifier model. A high classification performance was obtained for both images (kappa > 0.88, overall accuracy > 90%). After the classification process, LULC maps for both years were generated, and statistical calculations for the LULC change were computed for both the entire study area (15 × 25 km) and a buffer zone with a radius of 1 km around the power plant. In the whole study area, artificial surfaces significantly increased (78.46%), whereas forests (- 8.31%) and barren lands experienced a considerable decrease (- 6.11%). In the 1 km buffer, artificial surfaces predominantly increased (113.94%), while forests and barren lands decreased dramatically (- 69.13% and - 74.28%, respectively). The agricultural areas in the study area were changed into other LULC classes: 9.1% to artificial surfaces, 27.6% to barren lands, and 21.7% to forest. The rise in the area of artificial surfaces was especially noticeable within the 1 km buffer zone: construction activities converted 36.1% of agricultural fields, 54.1% of forests, and 23.2% of barren lands into artificial surfaces. The filling activities on the seashore resulted in a loss of water bodies of up to 26.5%. The study provides an overview of how the LULC classes have evolved on the construction site and in the region. In the end, the study discusses how the current land use preferences in the region contradict the issues and concerns mentioned in the existing body of literature.