A method for evaluating ecological quality levels and trends using natural evolution time series.

Ecological quality assessment serves as a valuable tool for impartially measuring the effects of development and utilization activities on ecosystems, forming the bedrock of ecological governance. The ecological quality evaluation method, relying on a reference frame, takes into consideration the comparability of ecological quality, thus guaranteeing the presence of external benchmarks for the assessment results. However, present research on absolute assessment of ecological quality is limited, as the majority of existing studies focus on large-scale regions, and overlooking regional differences in smaller-scale regions to a certain extent. Therefore, to achieve a comprehensive, convenient, objective, and accurate ecological quality evaluation, this study established a model for regional ecological quality assessment through remote sensing. This model was based on the analysis of natural evolutionary features and was applied to assess the ecological quality in the largest ecosystem within the arid area of the Ningxia Hui Autonomous Region (NHAR) - the grassland ecosystem. For this assessment, the Modified soil-adjusted vegetation index (MSAVI), the wetness components of tasseled cap transformations (WET), and net primary productivity (NPP) were selected as indicators of soil and water conservation capacity, water source conservation capacity, and carbon sequestration capacity, respectively. The evaluation units were established according to the geomorphic types, and the improved derivative dynamic time warping (DDTW) algorithm was used to compare the conformity between the time series of evaluation units and natural evolution time series, pixel by pixel. The ecological quality was classified into five levels (levels 1, 2, 3, 4, and 5) using the natural breaks (Jenks) classification method. The Mann-Kendall trend test and the Sen's slope are used to analyze the trend of the conformity measurement results of the ecological quality each year at a significance level of 95%. The main results are as follows: (1) The indicators selected in the study are mostly independent of each other within each evaluation unit, with a correlation (0.552) observed only between MSAVI and WET in plain areas. (2) Spatial heterogeneity exists in the natural evolution characteristics, with significant variations in the similarity between the reference frames of different evaluation units. There is no strict rule linking the similarity results to the elevation of the terrain. (3) From the perspective of the evolutionary process and intensity, the soil and water conservation capacity, water source conservation capacity, and carbon sequestration capacity in the study area primarily fell within level 2. The advantageous level of each evaluation unit was also primarily at level 2. The overall trend of these three capacities demonstrated stability and remained constant. In summary, our assessment indicated that the overall quality of grassland ecosystems in the arid areas of the NHAR was relatively stable and superior. This method we employed has broad applicability and can serve as a general framework for ecological quality assessments. It can be adapted for ecological restoration assessments in other regions by fine-tuning regional parameters, offering valuable insights for regional ecological management.