RESUMO
Lakes, as vital components of the Earth's ecosystem with crucial roles in global biogeochemical cycles, are experiencing pervasive and irreparable worldwide losses due to natural factors and intensive anthropogenic interferences. In this study, we investigated the long-term dynamic patterns of the Tonle Sap Lake, the largest freshwater lake in the Mekong River Basin, using a series of hydrological data and remote sensing images between 2000 and 2020. Our findings revealed a significant decline in the annual average water level of the lake by approximately 2.1 m over 20 years, accompanied by an annual average reduction in surface area of about 1400 km2. The Tonle Sap Lake exhibited episodic declines in water level and surface area, characterized by the absence of flooding during the flood season and increasing aridity during the dry season. Furthermore, the shoreline of the lake has significantly advanced towards the lake in the northwestern and southern regions during the dry season, primarily due to sedimentation-induced shallowing of the lake edge depth and decreased water levels. In contrast, lake shorelines in the eastern region remained relatively stable due to the constructed embankments for the protection of the cultivated farmland. While the seasonal fluctuations of the Tonle Sap Lake are regulated by regional precipitation in the Mekong River Basin, the prolonged shrinking of the lake can be mainly ascribed to intensive anthropogenic activities. The interception of dams along the upper Mekong River has resulted in a decrease in the inflow to Tonle Sap Lake, exacerbating its shrinkage. Moreover, there are minor impacts from agricultural land expansion and irrigation on the lake. This study highlights the driving forces behind the evolution of Tonle Sap Lake, providing valuable information for lake managers to develop strategies aimed at conserving and restoring the ecological integrity of the Tonle Sap Lake.
RESUMO
Mangrove forests provide vital ecosystem services. The increasing threats to mangrove forest extent and fragmentation can be monitored from space. Accurate spatially explicit quantification of key vegetation characteristics of mangroves, such as leaf area index (LAI), would further advance our monitoring efforts to assess ecosystem health and functioning. Here, we investigated the potential of radiative transfer models (RTM), combined with active learning (AL), to estimate LAI from Sentinel-2 spectral reflectance in the mangrove-dominated region of Ngoc Hien, Vietnam. We validated the retrieval of LAI estimates against in-situ measurements based on hemispherical photography and compared against red-edge NDVI and the Sentinel Application Platform (SNAP) biophysical processor. Our results highlight the performance of physics-based machine learning using Gaussian processes regression (GPR) in combination with AL for the estimation of mangrove LAI. Our AL-driven hybrid GPR model substantially outperformed SNAP (R2 = 0.77 and 0.44 respectively) as well as the red-edge NDVI approach. Comparing two canopy RTMs, the highest accuracy was achieved by PROSAIL (RMSE = 0.13 m2.m-2, NRMSE = 9.57%, MAE = 0.1 m2.m-2). The successful retrieval of mangrove LAI from Sentinel-2 can overcome extensive reliance on scarce in-situ measurements for training seen in other approaches and present a more scalable applicability by relying on the universal principles of physics in combination with uncertainty estimates. AL-based GPR models using RTM simulations allow us to adapt the genericity of RTMs to the peculiarities of distinct ecosystems such as mangrove forests with limited ancillary data. These findings bode potential for retrieving a wider range of vegetation variables to quantify large-scale mangrove ecosystem dynamics in space and time.
