En fjärranalytisk metod till säsongsförändringar: NDVI och snötäcketrender : Övervakning av miljöförändringar genom vegetations- och snöindex härledda från Landsat och MODIS

Understanding seasonal patterns of vegetation and snow is essential in studies of climate change and monitoring ecosystem responses. Previous research presents a shortening in the duration of snow cover along with an increase in spring temperature and precipitation leading to a longer growth period for vegetation. Satellite imagery provides frequent cover over geographically vast areas without the need for additional infrastructure or personnel. As such, this study uses satellite-derived values to investigate phenology trends and snow cover dynamics in central and southern Sweden over recent decades. Landsat data from 1985-2024 was used to measure vegetation greenness, through the Normalised Difference Vegetation Index (NDVI). Observations acquired by the Moderate Resolution Imaging Spectroradiometer (MODIS) from 2000 to 2024 measured snow cover through the Normalised Difference Snow Index (NDSI). Phenology and snow cover metrics were extracted on a monthly basis, through fixed thresholds, such as the Start of Season (SOS), End of Season (EOS) as well as the length of the winter season and Growing Season Length (GSL). A comparison between the extracted metrics and climate variables from the Swedish Meteorological and Hydrological Institute (SMHI) was conducted. Additionally, the extracted metrics were converted to days using a standardised month length of 30.43685 days to establish a changing rate of days per year, which were then compared to reference datasets on vegetation periods and snow-covered days concluded by SMHI. Results indicate a shift in vegetation dynamics due to a trend of earlier onset of spring vegetation as a result of earlier snowmelt in the winter. A rise of NDVI throughout the season was found as well as a trend of declining snow coverage in the study area, except for April which shows signs of an increase in fleeting snowfall. A moderately positive correlation between NDVI and a rise in temperature (r = 0.27) and more precipitation (r = 0.30) was found at the beginning of the season, suggesting vegetation activity increase. A strong correlation between snow cover and snow depth (r = 0.79) declining during the time period was found, indicating an earlier end of the winter season. Combined trends suggest a seasonal change of lengthening of the growing season with +3.5 days per year during the entire period from 1985 to 2024, and +1.7 days per year post-2000, as well as a shortening of the winter by –0.8 days per year (2000-2024). The findings correspond to existing papers and measures of temperature, precipitation, and snow depth; demonstrating the validity of remote sensing methods for the purpose of examining seasonal shifts.