关键词: 增强型植被指数, 极端气候指数, 空间异质性, 相关分析, 最优参数地理探测器

Abstract:

Liaoning Province is located in the arid-humid climate transition zone and has experienced a significant increase in extreme climate events in recent years, which poses a serious threat to the stability of its ecosystems. As a core component of terrestrial ecosystems, vegetation plays a critical role in maintaining ecological stability. Investigating vegetation dynamics and their responses to extreme climatic events in Liaoning Province can inform regional climate adaptation strategies and contribute to enhancing ecosystem resilience. However, the response of vegetation in this region to extreme climatic events remains poorly understood. The Enhanced Vegetation Index (EVI) is a widely used indicator in vegetation monitoring that can reduce the impact of the soil background and atmosphere, providing high sensitivity and superiority in monitoring vegetation change. The Expert Team for Climate Change Detection Monitoring and Indices (ETCCDMI) has established a set of extreme climate indices that systematically characterize the evolution of extreme climate events across multiple dimensions, including frequency, intensity, and duration. Studies have confirmed that the impact of extreme climate on vegetation has intensified in recent years, and the coupling relationship between climate and vegetation at the monthly scale is closer than that at the annual scale. However, most studies have focused on annual or seasonal scales, overlooking the dynamic responses of vegetation to extreme climate conditions at the monthly scale. Furthermore, traditional correlation analysis and linear regression are commonly used to explore the relationship between long-term vegetation and extreme climate; however, these methods have demonstrated limited capacity to capture the spatial heterogeneity of vegetation driven by nonlinear and interactive drivers. In contrast, the geographic detector effectively addresses this issue and can identify the dominant extreme climatic indices and their interactions that influence the spatial heterogeneity of vegetation. However, in previous studies, extreme climate indices were often processed by traditional geographic detectors using a single discretization method with empirically defined layer numbers, leading to systematic biases in the detection results. The optimal parameter geographic detector integrates multiple discretization methods and layer numbers for processing extreme climate indices, with the optimal combination of parameters selected based on the maximum explanatory power. This methodology facilitates the precise identification of vegetation spatial heterogeneity and its climatic drivers through the systematic elimination of the subjective bias inherent in traditional approaches. This study was based on monthly EVI and meteorological data (daily maximum temperature, daily minimum temperature, and daily precipitation), and RClimDex 1.0 was used to extract monthly extreme climate indices. A simple linear regression was applied to analyze the dynamics of changes in vegetation and climatic extremes in Liaoning Province from 2000 to 2020. Pearson’s correlation coefficients were used to explore the relationships between EVI and extreme climate indices, and the optimal parameter geographical detector was employed to identify the dominant extreme climate drivers of EVI and quantify their interactive effects on vegetation heterogeneity. The results indicated that 1) the 21-year mean EVI in Liaoning Province was 0.58, exhibiting an upward trend with a growth rate of 0.035 per decade. Spatially, the EVI exhibited a decreasing gradient from the northeast to the southwest. The area of high vegetation cover expanded by 32.7%, whereas the areas of medium and low vegetation cover contracted by 20.1% and 12.2%, respectively. Vegetation improvement was observed in 86.3% of the study area, with the most pronounced enhancements recorded in the forested zones of eastern Liaoning Province. Conversely, degradation was primarily concentrated in urban areas, riparian corridors, transitional belts from grasslands in western Liaoning to croplands in the central region, and the foothills of southeastern Liaoning, with vegetation degradation being the most significant in urban areas. 2) Extreme high-temperature and precipitation events have increased, whereas extreme low-temperature events have decreased in frequency. Among the extreme temperature indices, except for cold days (TX10P) and cold nights (TN10P), which showed slight declines of −1.537 and −0.367 days per decade, respectively, all remaining extreme temperature indices exhibited upward trends. Only TX10P and the minimum value of the daily minimum temperature (TNn) exhibited statistically significant trends (p<0.05). In addition, all extreme precipitation indices showed significant increasing trends (p<0.05), with larger magnitudes of change than those of extreme temperatures. The maximum 5-day precipitation amount (RX5day) increased the most at a rate of 14.492 mm/10a in the summer. 3) Among the extreme temperature indices, TX10P, TN10P, and diurnal temperature range (DTR) were predominantly negatively correlated with EVI, whereas all other indices were positively correlated with EVI. Regional heterogeneity in correlation strength was evident, with TNn and the maximum value of daily maximum temperature (TXx) being significantly positively correlated with EVI (p<0.05), with TNn showing the strongest correlation (r=0.957). A higher correlation between TNn and EVI was observed in the croplands of eastern Liaoning, as well as in the western and central regions, whereas TXx exhibited an inverse spatial pattern of correlation. Warm days (TX90P) and warm nights (TN90P) were insignificantly positively correlated with EVI in the croplands and forests of central-eastern Liaoning, in contrast to the insignificant negative correlations in the western grasslands. Conversely, TX10P and TN10P displayed reverse correlations. Extreme precipitation indices were significantly positively correlated with EVI (r=0.701), and their spatial distribution was characterized by an east-low to west-high gradient of precipitation. 4) Extreme precipitation had a greater effect on the spatial heterogeneity of EVI in Liaoning Province than extreme temperature, with RX5day being identified as the dominant driver (q=0.28). Among the extreme temperature indices, TX10P had the strongest explanatory power for EVI (q=0.23), whereas TN90P and TN10P exhibited negligible contributions (q<0.1). The interaction between all extreme climate indices exceeded the effect of a single index, exhibiting either Bivariate or Nonlinear enhancement. The interaction between RX5day and TNn was identified as the strongest driver of EVI (q=0.57), followed by the interaction between RX5day and DTR (q=0.56). Among the extreme temperature indices, the interaction between TX10P, and DTR, and TNn had the highest explanatory power (q=0.50 and 0.48, respectively), although it was lower than the interaction of extreme precipitation. This study innovatively combined correlation analysis and the optimal parameter geographic detector at a monthly scale, effectively avoiding the subjectivity inherent in traditional methods when analyzing nonlinear vegetation responses to extreme climate conditions. These findings provide valuable insights for formulating climate adaptation strategies and vegetation restoration frameworks in Liaoning Province, to enhance regional ecological resilience and advance environmental sustainability.

Key words: enhanced vegetation index, extreme climate indices, spatial heterogeneity, correlation analysis, optimal parameter geographic detector