Regulating sensitivity analysis of spatial patterns of high-density urban riverfront areas to the river cooling effect: a case study of the Grand Canal in Hangzhou, China

Abstract River blue-green spaces are an important type of urban cooling island, and their cooling effect is largely influenced by the spatial pattern of urban riverfront areas. These areas are the primary source of the urban heat island effect in high-density areas along urban rivers. However, the sensitivity of spatial pattern variables to the river cooling effect is not well understood. To address this question, the Grand Canal in the main city of Hangzhou, was selected as study area. This study identified three building space form variables and seven green space form variables based on area, spatial pattern, and correlated these indicators with land surface temperature using the Autogluon model, then calculated the river cooling effect from the land surface temperature, and tested the sensitivity of these spatial pattern indicators by comparing their effects on the river cooling effect. The results are that: (1) the Autogluon model accurately establishes the relationship between land surface temperature and spatial pattern indicators; (2) the river cooling index (RCI) of the Grand Canal in the main city of Hangzhou ranges from 0.67-5.35℃, with an average RCI of 2.62℃, and the river cooling distance (RCD) ranges from 25.6-370.4m, with an average RCD of 167.3m; (3) Building patch density, Overall greenspace patch cohesion index, and greenspace patch density exhibits high sensitivity to regulating the river cooling effect(RCE). This study provides valuable references for identifying key spatial pattern regulating indicators in urban riverfront spatial planning, and for enhancing the cooling effect of medium-sized urban cooling islands and optimizing the outdoor thermal comfort of areas along rivers through rapid simulation evaluation methods.