Quantifying the Spatio-Temporal Variations and Impacts of Factors on Vegetation Water Use Efficiency Using STL Decomposition and Geodetector Method

Abstract

Water use efficiency of vegetation (WUE), the ratio of carbon gain to water loss, is a valid indicator to describe the photosynthetic carbon–water coupling relationship. Understanding how and why WUE changes are essential for regional ecological conservation. However, the impacts of various factors and their interactions on the spatial variation of WUE remain uncertain in the arid land of Northwest China. Here, we selected the Qilian Mountains (QM) and Hexi Corridor (HC) as the study areas. Supported by the Google Earth Engine, we explored the spatio-temporal variations of WUE in QM and HC for 2002–2021 using STL decomposition (a seasonal-trend decomposition procedure), trend analysis, and the Hurst index. Then, the Geodetector method was applied to quantify impacts of geographical and eco-meteorological factors on the spatial variation of WUE. The WUE in HC was higher than that in QM. Interestingly, the opposite longitude zonality characteristics are shown in the QM and HC. In QM, the WUE showed an upward trend with longitude increasing, while a downward trend with longitude increasing in the oases of HC. The WUE of cropland was the highest (1.15 ± 0.35 gC kg−1 H2O), and that of alpine vegetation was the lowest (0.2 ± 0.15 gC kg−1 H2O). WUE showed a decreasing trend across the study area, almost certainly due to a drop from May to July during 2002–2021. The air temperature is the dominant factor influencing the spatial variation of WUE. In addition, the interaction of any two factors is greater than the independent influence of either factor alone. The Geodetector method proved to be effective for quantifying the impact of complex multi-factors on the spatial variation of WUE. This study provides a new technical scheme to analyze the spatio-temporal pattern and quantify the impact of multi-factors on the spatial variation of WUE. These findings aid in understanding underlying mechanisms of WUE variation and thereby will be beneficial for clarifying the response of vegetation to climate change.

Industrial Heat Source-Related PM2.5 Concentration Estimates and Analysis Using New Three-Stage Model in the Beijing–Tianjin–Hebei Region

“…STL decomposition is a non-parametric statistical method that consists of an inner loop for fitting trend and seasonal components and an outer loop for calculating robustness weights to reduce the impact of outliers [53]. Compared to other time-series decomposition methods, it offers greater precision and configurability.…”

Changes and driving forces analysis of alpine wetlands in the first meander of the Yellow River based on long-term time series remote sensing data

“…The primary objective of the interaction detection was to determine if the driving forces and the interaction's intensity interacted, i.e., the change in the explanatory power of the variable Y after the combination of different factors X. The principle was to calculate the values of q(X 1 ) and q(X 2 ) separately from q(X 1 ∩X 2 ) and determine the mode of interaction by comparing the magnitude of the values (Wang et al, 2022). The interaction relationship among the driving factors is shown in Table 4.…”

Spatio-Temporal Variability Analysis of Vegetation Dynamics in China from 2000 to 2022 Based on Leaf Area Index: A Multi-Temporal Image Classification Perspective

“…The STL method decomposes a time series (Y t ) into trend (T t ), season (S t ), and remainder (R t ) components. This filtering process reduces the impact of outliers and missing values on the trend and seasonal components [75,76].…”