Earth fissure hazard prediction using machine learning models.

Earth fissures are the cracks on the surface of the earth mainly formed in the arid and the semi-arid basins. The excessive withdrawal of groundwater, as well as the other underground natural resources, has been introduced as the significant causing of land subsidence and potentially, the earth fissuring. Fissuring is rapidly turning into the nations' major disasters which are responsible for significant economic, social, and environmental damages with devastating consequences. Modeling the earth fissure hazard is particularly important for identifying the vulnerable groundwater areas for the informed water management, and effectively enforce the groundwater recharge policies toward the sustainable conservation plans to preserve existing groundwater resources. Modeling the formation of earth fissures and ultimately prediction of the hazardous areas has been greatly challenged due to the complexity, and the multidisciplinary involved to predict the earth fissures. This paper aims at proposing novel machine learning models for prediction of earth fissuring hazards. The Simulated annealing feature selection (SAFS) method was applied to identify key features, and the generalized linear model (GLM), multivariate adaptive regression splines (MARS), classification and regression tree (CART), random forest (RF), and support vector machine (SVM) have been used for the first time to build the prediction models. Results indicated that all the models had good accuracy (>86%) and precision (>81%) in the prediction of the earth fissure hazard. The GLM model (as a linear model) had the lowest performance, while the RF model was the best model in the modeling process. Sensitivity analysis indicated that the hazardous class in the study area was mainly related to low elevations with characteristics of high groundwater withdrawal, drop in groundwater level, high well density, high road density, low precipitation, and Quaternary sediments distribution.

Probability assessment of vegetation vulnerability to drought based on remote sensing data.

Drought is one of the important factors causing vegetation degradation. Determination of areas with vegetation more sensitive to drought can be effective in drought risk management. Considering the ability to describe vegetation conditions, vegetation health index (VHI) was used to determine the probability of vegetation vulnerability to drought and to provide the map of Iran showing sensitive areas to drought. This study tries to express the probability of vegetation vulnerability to drought in four main climatic classes including hyper-arid, arid, semi-arid and semi-humid, and humid in Iran. Temperature condition index (TCI) and vegetation condition index (VCI) were calculated using land surface temperature (LST) derived from the MOD11A2 product and normalized different vegetation index (NDVI) obtained from MOD13A2 product, MODIS sensor. Combining these two indices, VHI was calculated for late of March, April, May, and June during 2000-2017. VHI was classified into five classes representing the drought intensity. Then, the probability of occurrence (%) of each class was calculated and multiplied with weight of each class, varying from 0 to 40 based on drought intensity. Finally, probability of vegetation vulnerability index (PVVI) was calculated by summing of the values obtained for each class. The results showed that PVVI was higher in arid and hyper-arid areas than that in other areas in the four studied periods. The highest mean values of PVVI in humid as well as semi-arid and semi-humid classes were found in April as 59.87 and 62.4, respectively, while the highest mean values of PVVI in arid and hyper-arid classes were observed in May as 70.98 and 68.13, respectively. In total, our results showed that PVVI is affected by different climatic and topographic conditions, and it suggested that this index be used to determine the probability of vegetation vulnerability.