2:15 PM - 2:30 PM
[ACG43-14] Spatio-temporal variability of vegetation response to climate in the forest-tundra ecotone of northeastern Siberia
Keywords:forest-tundra ecotone, NDVI, Arctic greening
Northern hemisphere terrestrial greening (positive trends in vegetation productivity) along with amplifying warming has been detected via field measurements and satellite observation over the past 40 years. Meanwhile, in some regions, a slowing or reversing response to warming trends has been reported in the recent decade. Understanding spatial and temporal heterogeneity in vegetation response on climate variability, which depends on differences in plant functional type or regional environment, is critical to studying vegetation dynamics over the northern high latitude under climate change.
This study investigated spatio-temporal variability of vegetation response in the forest-tundra ecotone region of northeastern Siberia to reveal 1) spatial variability in greening/browning trends, 2) regulation factors shift from temperature to water, and 3) perennial accumulated effect of land wetness condition on vegetation activity. We analyzed spatial variation of satellite-derived vegetation index and its correlation to climate variables in the last two decades from 2001 to 2020. Our study area covers the arctic coastal lowland to the mountainous region (over 2000 m) underlying continuous permafrost in northeast Siberia (60–75° N and 110–170° E). Vegetation also shifts from tundra to shrub and taiga along latitude and altitude. The annual maximum of the normalized difference vegetation index (hereafter, NDVImax) was derived from Moderate Resolution Imaging Spectrometer (MOD13A2) data. Climate variables such as thawing index (TI, annual sum of daily air temperature above 0°C), freezing index (FI, similar to TI but below 0°C in a year starting from previous September) and Climate Moisture Index (CMI, difference between annual precipitation and potential evapotranspiration) calculated from climate reanalysis data of ERA5-Land are prepared.
We examined 20-year time series of NDVImax, which were aggregated to 1° grid, by K-means cluster analysis, and divided them into four representative clusters; no clear trend region along the arctic sea, increase trend region east to south part, and the rest two intermediate clusters were scattered over the study area. The correlation of NDVImax and climate variables calculated for each pixel (0.1° resolution) showed heterogeneous spatial distribution. Positive correlation with the TI concentrated in the higher latitude area, while the positive correlation with FI expanded over the entire study area. The area with the significant correlation with FI shifted to the southeast from the first decade (2000-2010) to the latter (2011-2020). Vegetation response to climate is heterogeneous temporally and spatially, but we indicate climate parameters for seasons when vegetation doesn't grow may be important for vegetation. We calculated CMI for a year (from the previous September to the current year August) and two and three years (from two and three autumns ago). A positive correlation of NDVImax and 1-year CMI was mainly found around the northeast part, and its distribution area reduced in the recent decade. Toward the southeast, a positive correlation with perennial CMIs was pronounced rather than with 1-year CMI. This result indicates the possibility that precipitation allowed vegetation to grow after two years via water storage in the upper permafrost.
This study investigated spatio-temporal variability of vegetation response in the forest-tundra ecotone region of northeastern Siberia to reveal 1) spatial variability in greening/browning trends, 2) regulation factors shift from temperature to water, and 3) perennial accumulated effect of land wetness condition on vegetation activity. We analyzed spatial variation of satellite-derived vegetation index and its correlation to climate variables in the last two decades from 2001 to 2020. Our study area covers the arctic coastal lowland to the mountainous region (over 2000 m) underlying continuous permafrost in northeast Siberia (60–75° N and 110–170° E). Vegetation also shifts from tundra to shrub and taiga along latitude and altitude. The annual maximum of the normalized difference vegetation index (hereafter, NDVImax) was derived from Moderate Resolution Imaging Spectrometer (MOD13A2) data. Climate variables such as thawing index (TI, annual sum of daily air temperature above 0°C), freezing index (FI, similar to TI but below 0°C in a year starting from previous September) and Climate Moisture Index (CMI, difference between annual precipitation and potential evapotranspiration) calculated from climate reanalysis data of ERA5-Land are prepared.
We examined 20-year time series of NDVImax, which were aggregated to 1° grid, by K-means cluster analysis, and divided them into four representative clusters; no clear trend region along the arctic sea, increase trend region east to south part, and the rest two intermediate clusters were scattered over the study area. The correlation of NDVImax and climate variables calculated for each pixel (0.1° resolution) showed heterogeneous spatial distribution. Positive correlation with the TI concentrated in the higher latitude area, while the positive correlation with FI expanded over the entire study area. The area with the significant correlation with FI shifted to the southeast from the first decade (2000-2010) to the latter (2011-2020). Vegetation response to climate is heterogeneous temporally and spatially, but we indicate climate parameters for seasons when vegetation doesn't grow may be important for vegetation. We calculated CMI for a year (from the previous September to the current year August) and two and three years (from two and three autumns ago). A positive correlation of NDVImax and 1-year CMI was mainly found around the northeast part, and its distribution area reduced in the recent decade. Toward the southeast, a positive correlation with perennial CMIs was pronounced rather than with 1-year CMI. This result indicates the possibility that precipitation allowed vegetation to grow after two years via water storage in the upper permafrost.