单个遗址区古土地利用范围预测模型的构建——以河南仰韶村遗址为例  

Predictive Model of Area of Ancient Land-Use in One Site-A Case Study of Yangshao Village Site, Henan Province

作　　者：查理思[1,2] 吴克宁[1,2] 冯力威 魏洪斌[1,2] 

机构地区：[1]中国地质大学(北京)土地科学技术学院,北京100083 [2]国土资源部土地整治重点实验室,北京100035

出　　处：《土壤通报》2016年第5期1065-1070,共6页Chinese Journal of Soil Science

基　　金：国家自然科学基金项目(41371226)资助

摘　　要：以河南仰韶村文化遗址为研究对象,选择海拔高度、坡度、坡向和距河流水平距离为自变量,以遗迹点和非遗迹点为因变量,采用二分变量逻辑回归法进行分析,建立古土地利用范围预测模型。结果显示古土地利用分布概率与海拔高度成正比,与坡度值、坡向值和距河流水平距离成反比,其中坡度和坡向的影响程度最为显著,这表明古人类选择土地利用范围时,优先考虑坡度和坡向。基于预测模型,生成古土地利用分布概率等级图,结果显示南部和东部概率高,这表明古人类选择土地利用范围时,优先考虑南部和东部。通过实地调查,并结合考古资料,发现南部地势平坦、引水便利,适合农业生产和居住,已在该区域发现大量遗迹;东部虽地势平坦,但取水不便,适合其他活动,已在该区域发现埋葬。预测结果与实际结果较为吻合,表明模型具有一定的科学性。This paper took the archaeological remains of Yangshao Village site as the research object. The environmental indexes including altitude, slope, aspect, the distance to the river were selected as independent variables, while the archaeological remains and un- archaeological remains were selected as dependent variables. Then binary logistic regression was used to create a predictive model of areas of ancient land-use. The results showed that the probability of the distribution of ancient land use was proportionate to the value of altitude, was inversely proportional to the value of slope, aspect and the distance to the river. The influence of slope and aspect was the most significant. Based on the model, the grade map of the probability of distribution of ancient land-use was generated. The result showed that the probability in the south and east was high, which meant ancient human preferred in these areas. Combining field-survey and archaeological data, it was plain and available for agriculture in the south, and a lot of remains was found here. In the east, it was also plain but not available for agriculture, so other activities were taken here, such as burial. The predictive results scientificity. well with the actual results, which proved this model had certain scientificity.

关 键 词：仰韶村遗址 遗迹点 二分变量逻辑回归 预测模型 

分 类 号：K928.5[历史地理—人文地理学]