Enhancing vegetation formation classification: Integrating coarse-scale traditional mapping knowledge and advanced machine learning.

Mapping vegetation formation types in large areas is crucial for ecological and environmental studies. However, this is still challenging to distinguish similar vegetation formation types using existing predictive vegetation mapping methods, based on commonly used environmental variables and remote sensing spectral data, especially when there are not enough training samples. To solve this issue, we proposed a predictive vegetation mapping method by integrating an advanced machine learning algorithm and knowledge in an early coarse-scale vegetation map (VMK). First, we implemented classification using the random forest algorithm by integrating the early vegetation map as an auxiliary feature (VMF). Then, we determined the rationality of classified vegetation types and distinguished the confusing types, respectively, based on the knowledge of the spatial distributions and hierarchies of vegetation. Finally, we replaced each recognized unreasonable vegetation type with its corresponding reasonable vegetation type. We implemented the new method in upstream of the Yellow River based on GaoFen-1 satellite images and other environmental variables (i.e., topographical and climate variables). Results showed that the overall accuracy using the VMK method ranged from 67.7 % to 76.8 %, which was 10.9 % to 13.4 % and 3.2 % to 6.6 %, respectively, higher than that of the method without the early vegetation map (NVM) and the VMF method, based on cross-validation with 20 % to 60 % random training samples. The spatial details of the vegetation map using the VMK method were also more reasonable compared to the NVM and VMF methods. These results indicated that the VMK method can distinctly improve the mapping accuracy at the vegetation formation level by integrating knowledge of existing vegetation maps. The proposed method can largely reduce the requirements on the number of field samples, which is especially important for alpine mountains and arctic region, where collecting training samples is more difficult due to the harsh natural environment.

Improved soil surface nitrogen balance method for assessing nutrient use efficiency and potential environmental impacts within an alpine meadow dominated region.

The soil surface nitrogen balance (SSNB) method is commonly used to assess the nutrient use efficiency (NUE) of agricultural systems and any associated potential environmental impacts. However, the nitrogen flow of wide natural grasslands and other natural areas differ from that of artificial croplands and mown grasslands. In this study, we integrated root growth and the important nutrient resorption process into the SSNB model and used the improved model to clarify the nitrogen (N) flow and balance in the Three Rivers Headwater Region (TRHR)-an area dominated by alpine meadows-from 2012-2019. In the grassland system, the N surplus (ΔN) was 0.274 g m-2 year-1, and root return (BLD) dominated the N input, accounting for 67% of the total input (3.924 g m-2 year-1). N resorption was the main internal N flow in the grassland system (1.079 g m-2 year-1), and 30% of grassland uptake (NUP-grass). The ΔN of the agricultural system was 1.097 g m-2 year-1, which was four times that of the grassland, and chemical fertilizer was the largest input, accounting for 84% of the total input. The NUE in grassland was 93%, which suggests a risk of soil mining and degradation, while that of cropland was 76% and within an ideal range. The ΔN provides a robust measure of river N export, the TRHR was divided into three catchments, and the export coefficient was 16.14%-55.68%. The results of this study show that the improved SSNB model can be applied to a wide range of natural grasslands that have high root biomass and resorption characteristics.

Trends in total nitrogen concentrations in the Three Rivers Headwater Region.

It is important to protect the quality of the water in the Three Rivers Headwater Region (TRHR), known as the water tower of China, to guarantee the water security in downstream areas. However, because of a lack of long-term studies that span wide geographical areas, it is difficult to understand how the water resource in the TRHR should be protected. In this paper, we report the findings from our analysis of total nitrogen (TN) concentration data from 39 river monitoring stations for the period from 2012 to 2018. The water quality status was evaluated by comparing the concentrations with the national standards and calculating exceedance ratios for surface water. Trends were calculated with ordinary linear least-squares regression and a weighted least-squares (WLS) meta-analysis method. The results showed that the annual average TN concentrations in the TRHR rivers from 2012 to 2018 ranged from 0.68 to 1.06 mg/L, and were lower than those in the downstream reaches but higher than the global average in natural river waters. For the period from 2012 to 2018, the TN concentrations showed a highly significant increase (0.03 mg/L/year) across the entire TRHR and were increasing and decreasing at 71.8 % and 28.2 % of the stations, respectively. From the trend results, we divided the study area into two zones, one with increasing TN concentrations and one with decreasing TN concentrations. It is found that environmental factors had little influence on TN concentrations in the increasing and decreasing areas, but artificial factors such as population and restoration project areas contributed to the increases in TN concentrations in the increasing area. The TRHR remains a source of clean water in China; however, the water quality should be monitored closely, and measures should be implemented to protect the resource and mitigate the disturbances caused by human activities.