A comprehensive analysis of terrestrial surface features using remote sensing data

Abstract

Using the remote sensing data, this study aims to enhance our understanding of land surface features, including ecosystem distribution in association with topographic controls and climatic controls, vegetation disturbance due to natural hazards, and surface temperature changes with consideration of the influence of urbanization. In this study, the Global Inventory Monitoring and Modeling System (GIMMS) Normalized Difference Vegetation Index (NDVI) data sets from 1982 to 2006 were used to explore vegetation variation. A data mining method, Exhaustive Chi-squared Automatic Interaction Detector algorithm, was successfully applied to investigate the topographic influences on vegetation distribution in China. The study revealed that elevation is a predominant factor for controlling vegetation distribution among different topographic attributes (slope, aspect, Compound Topographic Index (CTI) and distance to the nearest river). Further, the study results indicated that solar radiation is the limited factor for plant growth in majority of the Northern Hemisphere in summer, and temperature is the main limitation for other seasons. Partial correlation coefficient (PCC) method was adopted to investigate the complex relationships of NDVI with weather variables (i.e., temperature, precipitation and solar radiation) and key climate indices (such as, El Niño-Southern Oscillation (ENSO), Indian Ocean Dipole (IOD), Arctic Oscillation (AO), and Antarctic Oscillation (AAO)). The study indicated that AO is the most significant index in affecting the temperatures in spring and winter in the Northern Hemisphere. This study enhanced the understanding of vegetation responds to asymmetric daytime (Tmax) and nighttime (Tmin) warming in different seasons. The result revealed that asymmetric warming of Tmax and Tmin may influence vegetation photosynthesis and respiration in the plant growth in different periods across biomes. In spring and autumn, vegetation in boreal and wet temperate regions of the Northern Hemisphere is positively correlated with Tmax and negatively correlated with Tmin, whereas, in dry regions the NDVI is always negatively correlated with Tmax and positively correlated with Tmin. In summer, the NDVI is negatively correlated with Tmax in many dry regions. In addition, this study developed a new index, Continued Vegetation Decrease Index (CVDI), to detect vegetation disturbance due to extreme natural hazards (such as, earthquake, wildfire, ice storms and so on). Using the Wenchuan earthquake occurred in Sichuan China on 12 May 2008 as an example, this study confirmed that the CVDI method can effectively identify the regions with severe vegetation damage, and it is expected that the newly-developed index can be used for detecting vegetation disturbance in other regions of the world. Finally, using the remote sensing data (land use data and surface temperature data) and weather station data, this study developed a new method to evaluate the urbanization influence on the temperature recorded at weather stations. The results revealed that the weather stations with most fast increase temperature are not in developed countries, but in developing countries. The results also imply that the global warming trend may be overestimated due to the under-estimation of urbanization influence on temperature increase.