jin Wang ^* Jiarui Han

• Sun Yat-sen University, Guangzhou, China

Step 1: Collation of historical literature and field investigation data. This aspect involves systematically collecting and analyzing materials from provincial, prefectural, and county gazetteers and other relevant local histories and official records to gather basic cultural and historical spatial information for the study area, emphasizing the impact of specific periods, events, or themes on the region. The collected historical documents and field survey data are then organized chronologically and compiled into an Excel database to form a comprehensive spatial-temporal information repository(Tuan, 1977, 1990). Step 2: Construction of a geographic database of historical settlements in Xinyi County (Blanco et al., 2019). Utilizing satellite maps and ArcGIS, the database contains coordinates (latitude and longitude) of topography, water systems, administrative divisions, prefectural and county seats, market towns, natural villages, and historical buildings. Following the logic of “backward tracing,” spatial information such as topography, water systems, administrative boundaries, prefectural and county seats, and settlements are reconstructed for various historical periods, moving from the present to the past. Step 3: Combining quantitative and qualitative analysis methods. First, based on the database in Step 2, spatial analyses were conducted using ArcGIS to calculate the kernel density index of the distribution of settlements and polders, the distance between settlements and water sources, and topographic changes to study the macro-distribution characteristics of settlements(Chainey et al., 2008; Sabel et al., 2005; Silverman, 2018). The formula for the kernel density analysis method is as follows： f(x,y)=1/(nh^2 ) ∑_(i=1)^n▒K (d_i/h) The equation represents the density estimate at the position, the number of observations, the smoothing parameter, the kernel function, and the distance from the position to the observation position. Data visualization techniques in ArcGIS and Excel were utilized to quantify the relationships between settlement distribution and natural geographic and historical-cultural factors. Furthermore, qualitative methods such as comparative studies and descriptive analysis are employed based on the spatial analysis and data visualization to transform the synchronic spatial structure of the settlements into a diachronic historical process, thereby revealing the driving mechanisms behind the spatial evolution of settlements in Xinyi County (Whitehand, 1977; Whitehand, 2019).