Remote Sensing Monitoring Of Grassland Vegetation Greenup And The Correlativity Between Greenup And Meteorological Factors

Global warming has caused changes in vegetation greenup. Exploring the relationship between grassland vegetation phenology and climate change, understanding the characteristics of spatiotemporal characteristics and its responses to climate, has important theoretical and practical significance to guide the grazing activities and to explore the role of the grassland ecosystem in the global carbon cycle. In order to reveal grasslands phenology`s spatiotemporal characteristics and its responses to climate in Xilingol league, the paper extract grasslands greenup from SPOT VGT NDVI filtered by D-L during 1999 to 2012 to analyze remote sensing greenup date years average, time standard deviation, and linear trend. Then calculate statistic relationships between Greenup date and temperature, precipitation on 8 stations, and spatial relationships between Greenup date`s trend and temperature, precipitation, temperature`s trend and precipitation`s trend, to explore the greenup date response to climate change. The main conclusions are as bellows:(1) Greenup date remote sensing method and precision verificationThis paper selects D-L filtering in NDVI time series reconstruction. D-L can not only smooth the NDVI curve but also not lose too much information. Then monitoring the grassland greenup through dynamic threshold method in Xilingol league. This paper verifies the accuracy through 162 sample plots of ground monitoring data during 2010 to 2012, the accuracy is 68%, the remote sensing greenup date`s RMSE is 8.7 days. It explains the dynamic threshold method can accurately extract the study area remote sensing greenup date. Using Xilinhot animal husbandry meteorological experiment station of Stipa krylovii and Leymus chinensis phenology data to verify the remote sensing greenup date, the accuracy of Stipa krylovii is 70%, the RMSE is 9 days; the accuracy of Leymus chinensis is 70%, the RMSE is 11 days.(2) Spatial-temporal pattern of grasslands greenup date in Xilingol leagueThe greenup date fluctuations of main counties during 1999~2012 are relatively large. The greenup date of Erenhot and Sonid Left Banner are relatively early. The greenup date of East Ujimqin Banner and West Ujimqin Banner are relatively early. The greenup date of warm desert grassland is early. The greenup date of warm meadow steppe grassland is late. The mean greenup date in Xilingol league is complex. Warm desert grassland greenup date mainly concentrated in early April and the middle of April, the standard deviation is 28.65 d, the volatility is bigger; warm steppe greenup date mainly concentrated in the middle of April and in late April, the standard deviation of 18.45 d; warm meadow steppe greenup date mainly concentrated in late April and early May, the standard deviation is 12.97 d, warm meadow steppe vegetation is the most stable. Linear trend of greenup date research shows that 57% of pixels are advanced, 43% of pixels are delayed, the average change trend is-1.5d/10 a, 8% of pixels` trend are advanced, 5.9% of pixels` trend are delayed. Linear trend increased from southeast to northwest, overall, the role of climate change on Xilingol league area grassland has promote greenup date in advance. The trend in Warm meadow steppe is small, about 80% of the pixels` trend are within ±10d/10 a, 43% of pixels` trend in warm grasslands and warm desert grasslands are greater than 10d/10 a.(3) The correlativity between greenup and meteorological factors in Xilingol leagueXilingol league area grasslands greenup date are significant negative correlation with the temperature and precipitation. Specifically, the average of greenup date and temperature of partial correlation coefficient is-0.18, 5 sites are negatively correlated, 3 sites are significant negative correlated. The average of greenup date and precipitation of partial correlation coefficient is-0.58, 8 sites are negatively correlated, 5 sites are significant negative correlated. Most of the sites, temperature and precipitation worked at the same time. The average regression coefficient between greenup date and temperature and precipitation is 0.54, 7 sites are significantly related. We can effectively simulate greenup date based on temperature and precipitation. Statistical results show that from the space, it has no significant correlation between greenup date trend and the precipitation in spring. But it has correlation between greenup date trend and the precipitation trend, greenup date trend and the temperature trend. Affected by some individual sites, the correlations have no significant negative correlation.Explore grasslands phenology`s spatiotemporal characteristics and its responses to climate in Xilingol league has important theoretical and practical value in enhancing China’s response to climate change, solving grassland protection and management in the presence of specific problems, providing guidance in local agricultural and animal husbandry production and pretecting vulnerable grassland ecosystem.