No abstract

Runoff and soil erosion on sloping farmland pose a serious threat to agricultural productivity. Soil and water conservation management is considered to be effective in controlling runoff and soil erosion, especially in sloped farmland. In this study, 1910 natural rainfall events were monitored to evaluate the effectiveness of soil and water conservation management and rainfall types on runoff and soil erosion. The results confirmed that fish‐scale pit plots reduced runoff and soil loss by 70.18 and 91.07%. Runoff and soil loss on sloped farmland and naturally sloped plots were significantly higher (p < .05) than the other four soil management interventions. “Intense” events were the main causes of runoff and soil loss production, but the influence of “intense” rainfall events on runoff and soil loss was weakened by narrow terraced farmland. About 10% of the largest rainfall events represented from 21 to 27% of the rainfall depth, from 35 to 54% of the total runoff, and from 71 to 88% of the total soil loss. Rainfall intensity plays a dominant role in soil erosion, while rainfall accumulation is the main influencing factor for runoff. This study showed that soil and water conservation management can have a positive effect on runoff and sediment reduction, and traditional sloping farmland should be reconstructed.

In this paper, the binary images of 100 kinds of leaves are used for leaf recognition. Firstly, we screen 35 important features and use the grey clustering analysis to establish the quantitative feature system of leaves. Then we use the gradient descent tree algorithm (GBDT) to select core features and use probabilistic neural network (PNN) to recognize and classify leaves, constructing a hybrid GBDT-PNN model. In the end, we obtain the classification results of leaves to evaluate model performance and the influence of core features on the model. The results show that the accuracy rate of GBDT-PNN model using 12 core features is 92.75%. And the accuracy rate with all 35 features is 93.5%. It illustrates that the model has great performance and core features have high influence on the model. By comparing with other commonly used deep learning algorithms and models, it is verified that the GBDT-PNN image recognition and classification model is effective and has high accuracy.

No abstract