Effects of landcover fine-scale patterns on neighborhood-level winter and summer nocturnal and diurnal air temperatures

Abstract

ContextThere is a gap of knowledge on the effects of fine resolution landcover patterns on the distribution of air temperatures within neighborhoods, as well as on how these effects may differ depending on temporal (i.e., summer and winter, diurnal and nocturnal), and spatial (i.e. extent) scales.Objectives(1) Evaluate the effects of compositional and configurational fine-scale landcover patterns on the spatial distribution of air temperatures within neighborhoods. (2) Determine differences between winter and summer seasons and diurnal and nocturnal periods. (3) Evaluate if these effects relate to the spatial extent used for the analysis.MethodsRelationships between four landscape metrics and air temperature within four contrasting neighborhoods located in Santiago. Landcover was classified in six classes (built up, barren, grass, evergreen, deciduous, woody) from 1.5m resolution satellite images and temperature acquired from dataloggers located within neighborhoods. Linear mixed models were used for testing the relationships at six spatial extents.ResultsLandcover composition and configuration influence temperatures within neighborhoods, but these effects can greatly differ depending on the season, time of the day and extent of analysis. Grass and evergreen trees show the highest effects on neighborhood´s temperatures among the six landcover classes. Grass reduces summer temperatures at smaller extents but may increase temperatures at larger extents. Evergreen trees play a major role during the winter season increasing coldest nocturnal temperatures at all the analyzed extents. These vegetation effects appear to be mostly associated with the average and largest size of their respective patches.ConclusionsFine-scale landcover patterns play a role in regulating temperatures within neighborhoods, but these effects depend on the season, time of the day and spatial extent. Researchers and decision makers must be aware that results obtained at a given scale cannot be directly translated to another scales.