Spatiotemporal variability of grassland vegetation cover in a catchment in Inner Mongolia, China, derived from MODIS data products

Abstract

Conditions, distribution and development of vegetation in semi arid regions are highly variable. In this study we detected the temporal and spatial variability of vegetation in the Xilin river catchment from 2000 to 2008 by analysis of Moderate Resolution Imaging Spectroradiometer (MODIS) data products of that period. The study is based on LAI (Leaf area index) and supported by NDVI (Normalized difference vegetation index) and LST (Land surface temperature) data with a spatial resolution of 1 km. The mean LAI of the catchment from 2000 to 2008 is 0.59. Precipitation data of the study period governs the conditions and distribution of vegetation in the catchment. In dry years, e.g. 2001 and 2005, LAI was clearly lower (0.52) compared to 2003 (LAI=0.72) which was a wet year. As precipitation generally decreases from south-east to north-west, LAI values decrease according to this gradient. The influence of heavy grazing in the vicinity of the Xilin river is obvious as LAI is low (0.4) in these areas. The high temporal variability of the LAI is displayed by its high mean CV (coefficient of variation) which is 48% for the observed years. The analysis of sample areas illustrates temporal and spatial differences in vegetation development within the catchment and shows a generally delayed growth start in the north-west of the catchment.

Spatial simulation of evapotranspiration of semi-arid Inner Mongolian grassland based on MODIS and eddy covariance data

“…The MODIS LAI data were also used to model P for each pixel. Schaffrath et al (2010) found a linear relationship between precipitation and mean LAI for the vegetation period in the Xilin river catchment (R 2 = 0.78). We used this relationship to estimate P from the mean LAI (doy 113 to doy 297) of each pixel (Fig.…”

“…Additionally, several micrometeorological towers collected data (cp. Ketzer et al 2008;Wang et al 2009);Schaffrath et al (2010) analysed the spatiotemporal variability of precipitation and vegetation in this particular area using MODIS data. They found the conditions in the study area generally represent the steppe of the catchment, and the course of LAI is driven by precipitation, which had an annual mean of 210 mm with a high coefficient of variation (36%) from 2000 to 2008.…”

“…The MOD11A2 data contain the mean surface temperature for an 8-day time period at daytime (Ts_TD, in the morning around 10 a.m.) and at night (Ts_TN, around 11 p.m.). The data preparation method and useful references relevant for the data are presented by Schaffrath et al (2010). Ta_min_mo was derived from the mean Ts_TN of the 16 pixel as the surface temperatures from doy 113 (23 April) to doy 280 (07 Oct.) 1 of 2005 to 2008 correlate well (R 2 = 0.92) with the measured 8-day mean Ta_min_m of the period (Fig.…”

Spatially explicit regionalization of airborne flux measurements using environmental response functions

“…This is likely a function of geomorphological properties such as aspect, slope and soil type, but also due to the large variability of convective rainfall events across the study area (e.g. Schaffrath et al, 2011). Following superimposition of the footprints over the land cover data, the spatial contributions of different surface properties to each flux observation can be quantified (Fig.…”

“…Besides Tibet, Inner Mongolia is China's most important province for grassland-based livestock production, and desertification due to overgrazing is a major problem for extensive areas (Meurer and Jiang, 2001). In addition, the land cover in the investigation area varies distinctly in space and time (Ketzer et al, 2008;Schaffrath et al, 2011). We postulate that airborne EC flux measurement is a promising tool to gain new insights into the spatial variability of heat and moisture exchange across the XRC.…”