Trends in vegetation changes over wetland areas in Denmark using remote sensing data

Land cover dynamics play a vital role in many scientific fields, such as natural resources management, environmental research, climate modeling, and soil biogeochemistry studies; thus, understanding the spatio-temporal land cover status is important to design and implement conservation measures. Remote sensing products provide relevant information regarding spatial and temporal changes on the earth’s surface, and recently, time series analyses based on satellite images, and spectral indices have become a new tool for accurate monitoring of the spatial trend, and land cover changes over large areas. This work aims to determine the trends of vegetation spectral response expressed as the Normalized Difference Vegetation Index (NDVI) over the period 2005 and 2018 and compare these trends with the land-use and cover changes between 2005 and 2018 in wetland areas across Denmark. Change detection methods between two years based on bi-temporal information may lead up to the detection of pseudo-changes, which hinders the land-use and cover monitoring process at different scales. We studied the potentiality of including NDVI temporal curves derived from a yearly time-series Landsat TM images (30-m spatial resolution) to obtain more accurate change detection results. We computed the NDVI temporal trends using pixel-wise Theil-Sen and Man-Kendall tests, then we explored the relationship between NDVI trends and the different land-use and cover change classes. We found a significant relationship between NDVI trends and changes in land use and cover. Changes from cropland to wetland and cropland to forest coincided with statistically significant (p≤0.05) negative NDVI, and positive NDVI trends, respectively. Changes from grasslands to permanent wetlands corresponded with statistically significant negative NDVI trends. The difference in vegetation productivity trends could be indicative of the combined effect of human activity and climate. We show that this combined analysis provides a more complete picture of the land use and cover changes in wetland areas over Denmark. This analysis could be improved if the NDVI time series is seasonally aggregated.