Research On Drought Monitoring And Terrestrial Ecosystem Response Based On Multi-Source Remote Sensing Data

In the context of global warming,drought seems to be occurring more frequently,which has negative impacts on agriculture,water resources,natural ecosystems,and society activities.Drought is considered to be one of the most complex but least understood natural hazards,and it involves dynamics within the atmosphere-vegetation-soil continuum.In the past decades,scientists from different disciplines have proposed various drought indices to monitor the drought,aiming to explore the cause and duration of drought and also to understand how drought affects the terrestrial ecosystems.The traditional drought index based on site meteorological data can effectively monitor the drought conditions surrounding the meteorological stations,but can not be applied to characterize and monitor the detailed spatial pattern of drought conditions at regional scale.The emergence and development of remote sensing technology make it possible to monitor the drought in large-scale areas.The availability of spatio-temporal information for a variety of variables related to atmosphere,vegetation and soil conditions acquired based on different satellite sensors and remote sensing inversion methods benefits comprehensively monitor the drought conditions.This paper combines multi-source remote sensing data to monitor the drought conditions and explore the response of ecosystems to the drought in different areas.Multi-source remote sensing data used here include site meteorological data,gridded reanalysis meteorological data,eddy covariance?EC?flux data,and satellite remote sensing based vegetation indices and canopy variables.By analyzing the dynamic changes of different variables during the drought,we can comprehensively monitor the regional drought conditions and explore the mechanism of ecosystem response to climate change.This dissertation can shed light on the analysis of climate-vegetation interactions,ecosystem functions and the global carbon cycle.The main work are summarized as follows:?1?We firstly evaluated the performance of the standardized precipitation evapotranspiration index?SPEI?in monitoring the wet and dry conditions in southwestern China.The SPEI,an improved indicator of standardized precipitation index?SPI?,incorporates temperature data for the calculation of potential evapotranspiration,and is more promising under the condition of global warming.In this study,monthly SPEI from1982 to 2012 was calculated using the precipitation and temperature data from 89meteorological stations in southwestern China,in order to provide a comprehensive analysis of the drought conditions.Based on the SPEI series for various time lags,the multi-scale patterns,the trend,and the spatio-temporal extent of drought were successively analyzed.In addition,two remote sensing based drought indices were used to explore whether the vegetation and soil moisture can respond to the SPEI.The results showed that SPEI can effectively monitor the drying trend of southwestern China and successfully captured the typical drought events occurring in recent years.The strong consistency between SPEI and two remote sensing based indices indicated that the SPEI can also characterize the drought conditions of vegetation and soil to a certain extent.?2?This paper also evaluated the feasibility of using the canopy water content?CWC?retrieved from the radiative transfer model to monitor the drought in a semi-arid grassland located in Qinghai Lake watershed.In our experiments,a widely used PROSAIL model combined with the Look-up Table?LUT?method were applied to retrieve the CWC.To improve the inversion efficiency,Sobol'sensitivity analysis method was used to decrease the dimensions of the“free”parameters in the PROSAIL model.We finally obtained multiple CWC maps in five selected years covering the different meteorological conditions for analyzing the response of CWC to drought.The results showed that the drought largely reduced the region-wide CWC values.It was explicit to distinguish the drought year from the wet and normal years in terms of the distributions of CWC values.Since the CWC relied on variations of both leaf area index?LAI?and equivalent water thickness?EWT?during the inversion process,it was considered to be more sensitive to the drought.The retrieved CWC had similar spatio-temporal patterns with the gross primary productivity?GPP?and evapotranspiration?ET?from Moderate Resolution Imaging Spectroradiometer?MODIS?,indicating that the retrieved CWC can also reflect the variations in GPP and ET.The performance of CWC in monitoring the drought of grasslands in a semi-arid area has been well demonstrated.?3?Solar-induced chlorophyll fluorescence?SIF?has been proved to closely relate to the photosynthesis of terrestrial vegetation.In the recent few years,satellite SIF measurements open a new perspective on the monitoring of vegetation photosynthesis at regional and global scales.This study provided the first evaluation on the relationship between SIF from the Orbiting Carbon Observatory-2?OCO-2?and GPP data from EC flux towers in temperate forests.The relationships between tower GPP and OCO-2 SIF at two retrieval bands?757 nm and 771 nm?and two timescales?OCO-2 overpass time and daily timescale?were examined.We also evaluated the relationships between tower GPP and MODIS vegetation indices to assess how OCO-2 SIF performs compared to these widely-used vegetation indices.We further investigated the correlations of OCO-2SIF with absorbed photosynthetic active radiation?APAR?and two environmental scaling factors representing for low temperature stress and water stress,respectively.Finally,we evaluated the performance of a non-linear model based on SIF using cross-validation,and also compared the SIF model against the MODIS GPP algorithm in GPP estimation.The results showed that SIF was effective in characterizing the photosynthesis of temperate forests.SIF contained both information from APAR and and environmental stresses.Our findings demonstrate the strong ability of chlorophyll fluorescence observed by OCO-2in estimating GPP for temperate forests and reveal its great potential in future ecosystem functioning and carbon cycling studies.?4?This dissertation also used satellite SIF data along with a variety of climate and vegetation datasets to examine the impacts of the 2015 drought on the photosynthesis of Amazon forests,and also to understand how solar radiation and precipitation jointly controlled forest photosynthesis during the severe drought.We analyzed the anomalies of the various vegetation and climate variables in 2015 relative to multiyear means?2007-2014?.Two important variables SIF normalized by PAR(SIF_PARnorm)and SIF_yield,derived from SIF when photosynthetically active radiation?PAR?and APAR were determined,could provide insight into the component processes that were responsible for the changes in SIF during the drought.Our results showed that in spite of severe dry conditions from August to December in 2015,the decrease of photosynthetic activity of the Amazon forests was not as high as expected.Althought the drought largely reduced the SIF_yield,concurrent higher PAR and the fraction of absorbed PAR?fPAR?partly neutralized the negative impact of the drought.The higher PAR resulted from lower cloud cover and higher fPAR could be explained by the variations of LAI.Our findings showed that the SIF integrates information from SIF_yield,fPAR,and PAR and is effective for diagnosing the photosynthesis status of Amazon forests.