Method: Part A
In this study the satellite images were retrieved by Landsat 5TM and by Modis sensors; more specifically Modis 11A1, which provides specific layers with daytime and nighttime maps. Landsat 5TM is a moderate resolution satellite (with a resolution of 30 m for all its bands, except for band 6 which has a resolution of 120 m resampled to 30 m), and which has a temporal frequency of 16 days. Modis, in turn has a resolution of 1 km, and screens the surface of the earth on a daily basis. The main difference between the imagery retrieved by these two satellites is the resolution (Landsat is much better than Modis), and the fact that Modis provides night-time land surface temperature maps, whereas with Landsat 5TM only daytime land surface temperature imagery can be obtained. The parameters used as UHI indicators are: storage heat flux (calculated using daytime Landsat imagery), night-time land surface temperature (which have been retrieved through Modis) or daytime land surface temperature (retrieved through Landsat 5TM).
In the urban environment, storage heat flux maps were used for small surface analysis, such as in the article “The Urban Heat Island Effect in Dutch City Centers: Identifying relevant indicators and first explorations” (Echevarria Icaza et al., 2016c) where the average hotspot size analysed is 70.6 ha, in turn, Modis night-time imagery was used for the analysis of larger areas such as in the study “Surface thermal analysis of North Brabant cities and neighbourhoods during heat waves” (Echevarria Icaza et al., 2016b) where the average city size analysed is 1,565 ha. In the study “Using satellite imagery analysis to redesign provincial parks for a better cooling effect on cities. The case study of South Holland” (Echevarria Icaza et al., 2016a) Modis images were also used for the analysis of small land use patch surfaces, and as a matter of fact insignificant variations were found between the different land uses, and this is probably due to the coarse resolution compared to the patch size. Daytime land surface temperature (retrieved with Landsat 5TM) was used because the storage heat flux is not an UHI indicator for water patches and for non-urban patches due to the evapotranspiration effect.
The set of UHI parameters analysed varied depending on the purpose of the study. In all studies NDVI and albedo were mapped, and depending on the nature of the investigation other specific parameters were added.