Characterising Three Decades Of Evolution Of Forest Landscape Pattern In Shanxi Province Based On Self-organizing Map From Landsat Time Series

The analysis of change processes in relation to spatial patterns is gaining increasing attention in landscape ecology.Forest landscape pattern change is a critical process in landscape dynamics,and has a significant impact on the biodiversity,habitat quality,and ultimately,variation in the overall function of a forest ecosystem.Forest landscape change is a prominent change process for land cover in the Shanxi province.The purpose of this study is to characterise the forest landscape pattern evolution over a 1*1 km~2 landscape units covering Shanxi.This study firstly preprocessed and generated the annual Landsat image with maximum clear pixels in Shanxi province,then we finished an annual land use classification scheme by Random Forest classifier in Google Earth Engine;Four landscape metrics,i.e.,aggregation index(AI),largest patch index(LPI),patch density(PD),and shape index(SI),were derived based upon annual forest distribution maps(1987–2018).We analyzed the temporal trend of these metrics using MK-test;and these metrics were used to identify three spatially heterogeneous pattern types:severely fragmentised pattern(SFP),incompletely fragmentised pattern(IFP),and relatively intact pattern(RIP),based on the integration of the self-organising map(SOM)with k-means algorithms.Subsequently,we described the change process of the forest landscape patterns and explored how land use change drives the forest landscape change.Results indicate that:(i)The selected metrics show a stable trend in most of areas from 1987 to 2018;but some landscape units exhibit uptrend in metrics’value.(ii)RIP is the dominant landscape pattern type;the forest with the SFP is more likely to disappear;80.7%landscape units experienced a pattern change,wherein the change times primarily concentrated between 2 and 4 and the change duration was primarily in the range of 7–17 years.A relatively high proportion of pattern change occurred for sequential periods(2002–2005).(iii)The change in area of cultivated land drive forest landscape pattern towards discrepant directions in different areas.We concluded that using multiple landscape metrics is a systematic way in measuring landscape pattern;landscape patterns can be identified effectively through the combined use of the SOM with the k-means cluster;The forest landscape pattern has its own internal dynamics that has a different effect on the change processes.