Causal relationship in the interaction between land cover change and underlying surface climate in the grassland ecosystems in China

Abstract

Land-climate interactions are driven by causal relations that are difficult to ascertain given the complexity and high dimensionality of the systems. Many methods of statistical and mechanistic models exist to identify and quantify the causality in such highly-interacting systems. Recent advances in remote sensing development allowed people to investigate the land-climate interaction with spatially and temporally continuous data. In this study, we present a new approach to measure how climatic factors interact with each other under land cover change. The quantification method is based on the correlation analysis of the different order derivatives, with the canonical mathematical definitions developed from the theories of system dynamics and practices of the macroscopic observations. We examined the causal relationship between the interacting variables on both spatial and temporal dimensions based on macroscopic observations of land cover change and surface climatic factors through a comparative study in the different grassland ecosystems of China. The results suggested that the interaction of land-climate could be used to explain the temporal lag effect in the comparison of the three grassland ecosystems. Significant spatial correlations between the vegetation and the climatic factors confirmed feedback mechanisms described in the theories of eco-climatology, while the uncertain temporal synchronicity reflects the causality among the key indicators. This has been rarely addressed before. Our research show that spatial correlations and the temporal synchronicity among key indicators of the land surface and climatic factors can be explained by a novel method of causality quantification using derivative analysis.

Urbanization-Driven Changes in Land-Climate Dynamics: A Case Study of Haihe River Basin, China

“…Because forests have higher NDVI compared to the other land covers (e.g., agricultural land, grassland/shrub) [39], we assumed that the forest area positively affected regional NDVI (pathway 4 : +). Previous studies suggested that forest had the effects of stabilization and buffering on regional LST, which diminishes LST change [27,39,40].…”

“…The principle of eco-climatology indicates that at the regional scale, the ecosystem succession accumulates its effects on the atmospheric system by both spatial and temporal dimensions through the cause-effect chain from basic land surface biophysical characteristics to the heat and hydrological cycle distribution environment, which was theorized as the land-climate interaction [15,27]. Following the eco-climatology principle, we first selected a set of relevant key variables including vegetation, land surface physical structure, aerodynamic properties, solar radiation, thermal, and hydrological cycles to analyze the land-climate compounding system of the study area.…”

“…Between the second and third layers, we constructed all potential links to quantify the interaction between land surface indicators (NDVI, albedo, and LST) and regional climate indicators (P and CWC). Among these links, we assumed that NDVI had the positive effects on P and CWC (pathway 9 and 10 : +) because the evapotranspiration of vegetation promotes the regional hydrological cycle [27,44]. In contrast, we assumed that LST had the negative effects on P and CWC (pathway 13 and 14 : −) by heating land surface and drying air water [15,39].…”

“…In the results of the land surface and climate indicators by different land use and cover types, we noted that variations of indicators on the bare land was significantly higher than on every other land cover (through paired T-test, p < 0.05), which is visually reflected in Figure 5. The bare land with very low vegetation coverage had a lower heat capacity and water retention ability, and so land surface energy and water environment on bare land is easier to alter through solar radiation and reflection [27,39]. That is why it has higher variability in the land surface and climate indicators than other land covers.…”

Detecting Land Degradation in Eastern China Grasslands with Time Series Segmentation and Residual Trend analysis (TSS-RESTREND) and GIMMS NDVI3g Data

“…These pixels indicated that there were significant structural changes in the ecosystem in the specified year. The similar results between RESTREND and TSS-RESTREND in Hulunbuir may result because NDVI changes cannot be derived from coarse remote sensing data [49]. Using finer resolution NDVI dataset may improve the accuracy of the land degradation analysis and include more spatial details.…”