摘要 Abstract：

Climate change is a critical global concern, necessitating rigorous analysis of long-term climate data to identify trends and patterns at regional level. Different geographic locations experience varying degrees of climatic variability and the impact of these changes can manifest differently over time. Spatiotemporal meteorological dynamics are crucial for assessing not only the overall trends but also the localized variations that can significantly affect specific regions. Climate variability in the Himalayan region is of paramount importance due to its unique geography and ecological significance. This study conducts a comprehensive spatiotemporal analysis of climate patterns over the South Sikkim, leveraging historical climate data spanning key variables such as temperature, precipitation, humidity, and relevant parameters. Gridded data from reputable sources, including the Indian Meteorological Department (IMD) and the National Aeronautics and Space Administration (NASA), form the foundation of this investigation, allowing for a comprehensive examination of annual and seasonal trends. . In this work, nine grid points were used to analyze the long-term spatiotemporal shift of climatic variables in Sikkim, India. The primary statistical tools employed for this analysis are the Mann–Kendall (MK) test facilitates the identification of trends by assessing the statistical significance of monotonic trends, while the Sen Slope’s estimator quantifies the rate of change over time. The Pettit test is employed to pinpoint significant shifts in the climate data, aiding in the identification of abrupt changes in the observed trends. The overall results of the Mann–Kendall test indicate a slight decreasing trend in rainfall and increasing trends in temperature and relative humidity. The seasonal trends are particularly interesting; rainfall decreases during the monsoon period but increases in the pre-monsoon and post-monsoon seasons, while both temperature and humidity show an increasing trend in all seasons. On a monthly basis, the Mann–Kendall test reveals a decreasing trend for the months of June (Z = -1.20), July (Z = −2.96), August (Z = −4.76), and September (Z = −0.34) at a 1% level of significance. Conversely, from February to May, as well as October and December, there is an increasing trend with a 1% significance level. Change point analysis was found in all variables both annually and seasonally in a range of years, including 2003, 2009, and so on. Seasonal variance in rainfall and annual fluctuation in other factors have significantly changed, according to the study's conclusions.

關鍵字 Keywords：Climate change, Long term analysis, Spatio-temporal, Sikkim

摘要 Abstract：

Nitrogen deposition poses a significant environmental challenge globally, with North and Northeast Asia being particularly susceptible due to rapid industrialization and urbanization. This study investigates the temporal trends and wet nitrogen deposition patterns in this region from 2005 to 2020. The motivation behind this study stems from the urgent need to understand and address the increasing levels of nitrogen deposition in North and Northeast Asia. Rapid industrialization, urbanization, and agricultural activities have led to heightened nitrogen emissions, which can have detrimental effects on air quality, ecosystems, and human health in the region.

Research Questions：

1. What are the temporal trends in nitrogenous compound concentrations over North and Northeast Asia from 2005 to 2020?

2. What are the main drivers of nitrogen deposition in the study area, and how do they impact air quality and environmental health?

Concentration data for NH4, NO, NO2, and NO3 have been collected and spatially mapped for the years 2005, 2008, 2011, 2014, 2017, and 2020 using TROPESS Chemical Reanalysis gridded data. Trend analysis techniques are employed to discern significant temporal trends in atmospheric nitrogenous compound concentrations and wet nitrogen deposition.

This study is expected to provide valuable insights into the spatiotemporal dynamics of nitrogen deposition in North and Northeast Asia. By identifying trends and patterns in nitrogen deposition, it will contribute to a better understanding of regional air quality and environmental management needs. Preliminary findings may reveal the impact of industrial emissions, transportation activities, and agricultural practices on nitrogen deposition trends. Additionally, the study may shed light on the role of anthropogenic activity in nitrogen deposition.

關鍵字 Keywords：Nitrogen deposition, TROPESS, Air quality, Environmental management

摘要 Abstract：

The Himalayan region, characterized by its complex topography and diverse climatic conditions, faces growing challenges from changing climate patterns, including an increased susceptibility to landslides. Landslide susceptibility mapping plays a crucial role in mitigating the risks associated with landslides, especially under the influence of changing climate conditions.  This study presents an effective approach to evaluate the impacts of changing climate on landslide susceptibility by employing a hybrid stacking model. The model integrates Support Vector Machine (SVM), Artificial Neural Network (ANN), and Gradient Boosting Decision Trees (GBDT), while accounting for spatial heterogeneity to enhance accuracy in landslide susceptibility mapping. The unique combination of SVM, ANN, and GBDT in the stacking model leverages their respective strengths, addressing the multifaceted relationships between landslide occurrences and environmental factors in the Himalayan region. In this study, fourteen contributing factors and 106 landslide samples were gathered and applied to feature selection. The dataset was randomly divided into two subsets, with 70% used for training the landslide models and the remaining 30% for validation. The models were evaluated using the receiver operating characteristic (ROC) curve and area under curve (AUC). Ensemble techniques enhance landslide prediction model that exhibits robust performance with AUC values of 0.86 and 0.80 for training and testing, respectively which provide a strong framework for assessing the model's sensitivity and specificity, essential for reliable landslide risk assessments.

關鍵字 Keywords：Landslide inventory, Landslide susceptibility, Hybrid model, South Sikkim Himalayas

摘要 Abstract：

Problem/motivation: As a sprawling archipelagic nation with diverse mountainous landscapes, Indonesia is prone to various types of landslides. In this country, however, the current research primarily emphasizes the development of regional-scale landslide zoning maps, often neglecting detailed observations of landslide characteristics at specific landscapes. Consequently, obtaining detailed information on how and where to reduce landslide hazards poses a persistent challenge for local governments to date.

Aim: This study aims to conduct detailed geomorphological mapping to investigate the nature of landslides occurring on the southern flank of the complex Sumbing volcanic landscape situated at the centre of the Java Islands.

Methods: A combination of high-resolution LiDAR Digital Terrain Model (DTM), Global Navigation Satellite System (GNSS) data, multi-year Unman Aerial Vehicle (UAV)-based imagery, and field survey was used to create a 1:2500-scale geomorphological map. The identified landslides were subsequently categorized according to their type, activity level, and origin. Additionally, morphometric analysis including slope and dimension measurements, was conducted to ascertain the physical attributes of each landslide type within the research area.

Result: The map illustrates that most of the landscape within the study area has been influenced by mass wasting processes, encompassing both historical occurrences with unknown timing (relict landslides) and those observed from 2015 to the present (recent landslides). The geomorphological study uncovered that landslides in the research site could be initiated by natural factors like river and gully erosion, which can reduce slope stability or human activities such as excavating slopes exceeding 5 meters within the relict scarp zones. Although smaller in scale compared to many natural landslides, landslides induced by human intervention frequently result in more substantial damages, including the destruction of houses and the endangerment of lives.

Contributions: Therefore, the control of channel runoff and the enforcement of regulations concerning slope-cutting activities are crucial for reducing landslide hazards. The process of constructing a detailed geomorphological map of such kind is crucial for formulating future disaster management strategies, particularly in areas characterized by complex volcanic landscapes.

關鍵詞 Keywords：Detailed geomorphological map, landslides investigation, LiDAR DTM, disaster management, complex volcanic landscape