Abstract: In this paper, a total of 183 plots distributed in Harbin urban region were selected, and 4 different forest types (affiliated forests, AF; roadside forests, RF; landscape forests, LF and ecological welfare forests, EF) were classified, and tree growthrelated parameters and environmental factors (inside and outside the forest) were measured simultaneously. The aims of this study were to find differences between the 4 forests types in shading, cooling and humidifying effects and their associations with the environmental and tree growth characteristics. We found that urban forests produced a 77%-90% reduction of sun lights (shading effect) and a 3%-6% increase in relative humidity (humidifying effect). The forests had a horizontal cooling effect of around 3 ℃ lower than open sites (outside forests). Soil cooling effect was 1-2 ℃ lower under canopy than in outside forests. The vertical cooling effect characterized as canopy temperature about 1 ℃ lower than under canopy temperature. Shading, cooling and humidifying effects significantly differed among the four forest types. The comprehensive scores for all these microclimate regulations demonstrated that LF was 19% higher than the other three forest types on average. Urban forests had much stronger regulating functions at high air temperature, low air humidity and sunny days. Moreover, four treegrowth factors (tree height, canopy area, clear bole height and diameter at breast height) had significant influences on temperature reduction and relative humidity increase, and the general patterns were that larger trees usually accompany with higher regulating function. Our findings highlight that different forest types have dramatically different ecological service functions. For the landscape forests, we should continue to maintain its high ecological service function and strengthen the management of the forest types with low environmental regulation ability.

Key words: stable isotope, local meteoric water line, below-cloud evaporation effect, Northwest China, temporal and spatial distribution