Spatio-temporal Monitoring Of Mangrove And Its Response Relationship With Hydrometeorological Factors

Mangrove coastal wetland monitoring and its interaction with hydrometeorology are of important research value in the field of ecosystems.In this paper,using the Beibu Gulf region as the study area and based on the Google Earth Engine platform,we proposed a threshold segmentation method to extract mangrove growth regions based on the Landsat series and Sentinel-2 time-series images from 1986-2021,and assessed the applicability of 21 spectral indices to the spatial and temporal changes of mangroves.Secondly,the Detect-MonitorDetect(DMP)framework constructed by Land Trendr,BFAST,Mann-Kendall and other change monitoring algorithms was used to explore the spatio-temporal dynamic changes of mangroves over 36 years.Finally,this paper utilized Fourier harmonic analysis to extract the phenological characteristics of mangroves,and established a regression model of mangrove phenology and hydrometeorology(HANTS-PLSR)based on partial least squares method to determine the response relationships of mangrove intra-annual phenological changes to hydrometeorological factors.The coupling coordination model was also used to quantitatively assess the degree of coupling coordination between mangrove area change and each hydrometeorological factor.The conclusions of the study are as follows:(1)The overall accuracy of mangrove extracted by the method of threshold segmentation for 1986-2021 in Beibu Gulf is 0.887,and the extraction accuracy of different mangrove concentration areas is higher than 0.850,indicating that the method can effectively map mangrove growth areas.The method can extract large patches of mangroves and use them for change monitoring.In addition,the traditional vegetation indices can better reflect the spatial and temporal change processes of mangroves than the mangrove-sensitive indices,NDVI and TCA are suitable for capturing the dynamic changes of mangroves,and the information of mangrove change features captured by other indices has bias or even error.(2)The DMP framework provided a comprehensive tool for monitoring the spatiotemporal changes of mangroves,which effectively assessed the changes or persistence of mangroves in the past,present,and future in the Beibu Gulf.The detection results of the framework in the time domain showed that mangrove growth presented a long-term change process of flatness,growth and stability,which clearly represented the hierarchical gradient expansion law of mangrove,and provided the knowledge of the study area for monitoring mangrove status in near real-time.The historical monitoring results show that the average annual expansion rate from 1988 to 2020 was 239.822ha/year,and the expansion intensity was 0.031.The mangrove expansion rate gradually increased from 1988 to 2010,reached a peak of 499.893 ha/year from 2011 to 2015,and began to decline from 2015 to 2020.(3)Mangrove vegetation began to grow from February to May,and entered a lush period from July to August,until it began to decline from January to February of the following year.The changes of physiological parameters of mangrove vegetation were affected by water temperature and salinity.Air temperature is an important driving factor of vegetation growth in mangrove wetland,and only the physiological parameters of aleuria sinensis phenology are negatively correlated with it.The increase of humidity was negatively correlated with the change of water content in mangrove canopy.The influence degree between mangrove area expansion and hydroclimate gradually tended to a benign coupling coordination development level from 2000 to 2021,and the coupling coordination degree reached the highest in 2012-2019,which was above 0.7 and 0.5,respectively.