YU Enxu, ZHANG Mingfang, XU Yali, SUN Pengsen, MENG Zuozhu
Development and application of a regional forest water conservation function assessment tool
[Background] Multiple types of forest changes including deforestation, insect infestation, fire, and afforestation coupled with forest growth and restoration occur at a regional scale, which lead to the restoration effects of forest and associated water conservation capacity accumulate over space and time. The Zagunao watershed has suffered from deforestation in the past decades, which has severely reduced its water conservation capacity. This paper focuses on the forest change dynamics and the corresponding restoration of the water conservation capacity in this watershed under a series of ecological protection projects. However, traditional assessment tools of forest water conservation function lack the description of the spatial-temporal cumulation of forest growth and associated water conservation capacity, and are featured with complex computation, which impedes the evaluation on the effects of forest restoration on water conservation in a rapid and accurate way.[Methods] To address this issue, we innovatively developed the Equivalent Recovery Area (ERA) model for describing the hydrological function recovery with forest changes after forest deforestation and assessing the spatial-temporal dynamics of cumulative forest recovery after forest logging. Then, through combining the forest restoration evaluation model (ERA) and classic forest water storage function evaluation method (comprehensive water storage capacity method), this study took advantage of ENVI/IDL, ArcGIS Engine/C#.Net GIS to develop a GIS-based tool for assessing regional forest changes and associated forest water conservation function dynamics. The tool named Regional Forest Water Conservation Function Assessment Model (RFWCFAM) supports to evaluate the spatial-temporal changes of forest landscapes and forest water conservation functions under natural and artificial recovery scenarios. Charts and spatial distribution maps of forest changes, water conservation capacity dynamics, and restoration measures layout generated, after inputting precipitation, vegetation/forest, soil, and forest restoration measures data in RFWCFAM.[Results]This tool was successfully applied in the Zagunao watershed. According to the assessment, the EHFR (equivalent hydrological function recovery) coefficient increased in study area with the implementation of natural and artificial restoration measures from 2010 to 2030. The canopy, litter, soil and total forest water conservation capacity of the Zagunao watershed were about 129.8, 25.96, 523.62, and 697.37 t/hm2, respectively, in 2010, and all layers showed consistent upward tendencies from 2010 to 2030 under both natural restoration and artificial restoration scenarios. The artificial restoration measures yielded better effect on forest water conservation function than natural restoration. By 2030, the forest water conservation capacity of artificial restoration scenario is expected to be about 5.19% higher than that of natural restoration scenario.[Conclusions] The RFWCFAM has addressed the limitations of the traditional water resource assessment method and has been successfully applied and validated in the Zagunao watershed. Additionally, the RFWCFAM can provide an effective assessment tool to forest and water managers for evaluating the potential effects of forest restoration on water conservation function, along with scientific supports for adaptive forest management and soil and water conservation project evaluation and management in restorations of forest ecosystem functions in China.
Automatic extraction of disturbed patches in production and construction projects based on deep learning
[Background] Rapid acquiring the disturbance range of production and construction projects is critical for improving the efficiency of soil and water conservation supervision and inspection, and for controlling man-made soil erosion. The plots for production and construction projects are numerous and scattered, but remote sensing technology can achieve timely and accurate acquisition of their spatial distribution. At present, the remote sensing interpretation of disturbed patches in production and construction projects is mainly based on human-computer interaction with less automatic extraction. The existing algorithms are highly specialized and low in regional applicability, which makes it difficult applying in engineering.[Methods] The central and eastern parts of Gansu province were used as the research area, and deep learning algorithms were applied to carry out research on automatic extraction of disturbance areas. Based on GF-1 and GF-6 satellite images, the disturbance characteristics of production and construction projects were summarized, and bare land and newly added construction land were determined as the characteristic features detecting in production and construction projects. Selected 1 156 disturbances totally for the patch samples, then the enhanced DeepLabV3+ model and the U-Net model were utilized to construct an automatic extraction model of disturbance patches. Based on automatic extraction, the minimum recognition area was set to reduce the interference of mini patches. The extraction effects of different models and different minimum recognition areas were quantitatively evaluated by indicators such as quantity recall rate, quantity accuracy rate, area recall rate, and area accuracy rate.[Results] 1) The DeepLabV3+ model had fewer erroneous image spots. The quantity accuracy rate was more than 15% than that by the U-Net model, and the area accuracy rate was slightly higher than the U-Net model. 2) In Chengguan district and Zhengning county, the DeepLabV3+ model had fewer missing image spots, and the quantity recall rate was more than 5% higher than that of the U-Net model. The quantity recall rate of the DeepLabV3+ model in Baiyin district was slightly lower than that of the U-Net model. 3) Large patches extraction of these two models was better. The area accuracy rate was higher than the quantity accuracy rate, and the area recall rate was higher than the quantity recall rate. 4) As the minimum recognition area increased, the area accuracy of the model increased and the area recall rate decreased. 5) When the minimum recognition area was 1 hm2, utilizing the DeepLabV3+ model generated the best extraction effect. The average quantity accuracy rate was 78.31%, the quantity recall rate was 61.69%, the area accuracy rate was 88.82%, and the area recall rate was 80.60%.[Conclusions] In summary, DeepLabV3+ model has better overall classification effect. After setting an appropriate minimum recognition area, it can balance the accuracy and recall of the automatic extraction results. At the same time, the model has good extraction results in different districts and counties, which indicates that the deep learning algorithm has strong regional applicability and can be applied to a wide range of remote sensing supervision work.
Ecosystem service dynamics in sandy land in Northeast China during the past 30 years
[Background] The sandy land in Northeast China (Horqin sandy land, Songnen sandy land and Hulunbuir sandy land) belongs to the transition zone from semi-arid and semi-humid area to semi-arid area. The desertification of sandy land is serious, thus it is urgent to carry out ecosystem services. Revealing the temporal and spatial evolution characteristics of ecosystem services in this region over a long period of time may provide a scientific basis for ecological restoration and soil and water conservation in sandy land.[Methods] Based on the InVEST model and using data such as land use data, vegetation cover, digital elevation models, soil quality, etc., the article evaluated the functions of soil conservation, water conservation, carbon storage and habitat quality of this area from 1990 to 2020.[Results] 1) The total amount of soil conservation in sandy land of Northeast China increased by 1.98×107 t/a, and the high-value areas were concentrated in the southwest of Horqin sandy land, the east and west of Songnen sandy land, and the east of Hulunbuir sandy land. 2) The water conservation showed an overall increasing trend with an increase rate of 2.88×107 mm/a. The high-value areas were concentrated in the southeast of Horqin sandy land, the northeast of Songnen sandy land and the east of Hulunbuir sandy land. 3) The carbon storage increased by 1.60×106 t/a. The high-value areas were mainly distributed in the northeast of Horqin sandy land, the east and west of Songnen sandy land and the east of Hulunbuir sandy land. 4) The average habitat quality index changed little overall and it showed a trend of first increasing and then decreasing, the high-value areas were concentrated in the east and west of Songnen sandy land and the east of Hulunbuir sandy land.[Conculsions] The four major ecosystem service functions of soil conservation, water conservation, carbon storage and habitat quality in sandy land in Northeast China showed an overall increasing trend, and the increasing areas showed a certain spatial aggregation. From 1990 to 2020, the land use types of sandy land in Northeast China were mainly cultivated land, grassland and forest land. The area of cultivated land, construction land, sandy land and other unused land is on the increasing trend, while the area of forest land, grassland and water area is on the decreasing trend. This form of land use change has important implications for the spatiotemporal evolution of ecosystem services.
