Search

Derived from Landsat imagery and capable of enhancing the contrast between surface water bodies and the background, water indices are widely used in surface water body extraction. Whether one index and its optimal threshold can maintain the best all year round is a question. At present, however, little research has considered the effect of time or conducted experiments with data from different months. To identify the best index for surface water body extraction, two regions in Jilin Province were selected for the case study and a comprehensive comparative analysis considering the imagery acquisition time was conducted. Ten water indices were selected and calculated based on the 30 m spatial resolution Landsat TM/OLI imagery acquired in 1999 and 2001 and 2019 and 2021 from May to October. The indices included the Modified Normalized Difference Water Index (NDWI3 and MNDWI), Automated Water Extraction Index (AWEI) for images with and without shadow, Multi-Band Water Index (MBWI), New Water Index (NWI), Water Ratio Index (WRI), Sentinel-2 Water Index (SWI) originally calculated based on the Sentinel-2 imagery, New Comprehensive Water Index (NCIWI), Index of Water Surfaces (IWS), and Enhanced Water Index (EWI). The OTSU algorism was adopted to adaptively determine the optimal segmentation threshold for each index and the indices were compared in terms of inter-class separability, threshold sensitivity, optimal threshold robustness, and water extraction accuracy. The result showed that NWI and EWI performed the best in different months and years, with the best water enhancement effect that could suppress background information, especially for the water-related land use types and cloud pollution. Their optimal segmentation thresholds throughout the study period were more stable than others, with the ranges of −0.342 to −0.038 and −0.539 to −0.223, respectively. Based on the optimal thresholds, they achieved overall accuracies of 0.952 to 0.981 and 0.964 to 0.981, commission errors of 0 to 28.2% and 0 to 7.7%, and omission errors of 0 to 15% and 0 to 8%, with a kappa coefficient above 0.8 indicating good extraction results. The results demonstrated the effectiveness of NWI and EWI combined with the OTSU algorithm in better monitoring surface water during different water periods and offering reliable results. Even though this study only focuses on the lakes within two regions, the indices have the potential for accurately monitoring the surface water over other regions. [ABSTRACT FROM AUTHOR]

The reservoir rocks in the volcanic strata of the Jilin oil field are characterized by great complexity and diversity in composition and structure of lithology. To enhance the rate of lithology identification in subsurface is very laborious. However, lithology identification is often ignored in quantitative studies, though it is the basis for reservoir characterization. In this paper, an ensemble learning algorithm named gradient boosting decision tree (GBDT) was used to establish the classification model for the volcanic lithology identification of the Lower Cretaceous Yingcheng Formation in the Songliao Basin, NE China. At the same time, support vector machine (SVM), logistic regression (LR) and decision tree (DT) classification models were also adopted in contrast with the classification accuracy of GBDT model. Subsequently, the optimal key parameters for each model were determined by employing validation curves and GridSearchCv. These results indicate that the GBDT model is superior to the single classifier and can accurately distinguish the lithologic interface of breccia tuff and rhyolite. Moreover, it also has better recognition ability for thin layer. It was concluded that the ensemble learning algorithm GBDT has significantly enhanced the accuracy of lithology identification and can be used as a lithologic identification technology. • The GBDT model is applied well in volcanic lithology identification. • Parameter optimization is conducted to optimize the classifier. • The GBDT model is superior to the single classifier. • The GBDT model and has better recognition ability for thin layer. [ABSTRACT FROM AUTHOR]

Identifying ecologically vulnerable areas and understanding the responses of phenology to negative changes in vegetation growth are important bases for ecological restoration. However, identifying ecologically vulnerable areas is difficult because it requires high spatial resolution and dense temporal resolution data over a long time period. In this study, a novel method is presented to identify ecologically vulnerable areas based on the normalized difference vegetation index (NDVI) time series from MOD09A1. Here, ecologically vulnerable areas are defined as those that experienced negative changes frequently and greatly in vegetation growth after the disturbances during 2000–2018. The number and magnitude of negative changes detected by the Breaks for Additive Season and Trend (BFAST) algorithm based on the NDVI time-series data were combined to identify ecologically vulnerable areas. TIMESAT was then used to extract the phenology metrics from an NDVI time series dataset to characterize the vegetation responses due to the abrupt negative changes detected by the BFAST algorithm. Focus was given to Jilin Province, a region of China known to be ecologically vulnerable because of frequent drought. The results showed that 13.52% of the study area, mostly in Jilin Province, is ecologically vulnerable. The vulnerability of trees is the lowest, while that of sparse vegetation is the highest. The response of phenology is such that the relative amount of vegetation biomass and length of the growing period were decreased by negative changes in growth for dense vegetation types. The present research results will be useful for the protection of environments being disturbed by regional ecological restoration. [ABSTRACT FROM AUTHOR]