论文部分内容阅读
使用历史文献、气象、自然地理资料、数字高程模型和GIS技术,利用Logistic回归模型,分析兰州市地形要素对泥石流灾害发生的影响.结果显示:近百年兰州市发生泥石流灾害的时间比较频繁、范围较小;利用Logistic回归建立泥石流灾害预测模型得出发生泥石流灾害的概率与坡度、坡向相关性不显著,而与高程、距河距离成负相关关系,对模型进行检验,预测正确率达73.7%,预测效果较好.
Using the historical documents, meteorological data, natural geographic data, digital elevation model and GIS technology, the author uses Logistic regression model to analyze the impact of landform elements on the occurrence of debris flow disaster in Lanzhou City. The results show that the landslide occurred frequently in Lanzhou City in recent 100 years, And the Logistic regression model was used to establish the debris flow disaster prediction model. The correlation between the probability of occurrence of debris flow disaster and slope and slope aspect was not significant, but was negatively correlated with the elevation and distance from the river. The prediction accuracy of the model was 73.7 %, The forecast effect is better.