Correlation of Land Surface Temperature with IR Albedo for the Analysis of Urban Heat Island

: Albedo and Land Surface Temperature (LST) are thermophysical parameters that deﬁne the behavior of cities in terms of Urban Heat Islands (UHIs). Both parameters are correlated in such a way that materials with low values of albedo (associated with low reﬂection rates of solar radiation) result in higher heat absorption, and consequently, in higher LST values. This tendency reinforces the effect of UHI. Thus, the use of materials with high values of albedo in building envelopes can be a solution to reduce heat accumulation within cities and to subsequently improve the temperature reduction at nighttime.


Introduction
The increase in global population and the fast urbanization and industrialization concentrated in big cities have provoked a great revolution in land use worldwide. This revolution is accompanied by an important thermal change, creating a thermal disequilibrium between urban and rural areas. This thermal disequilibrium is known as the Urban Heat Island (UHI) effect, where heat is accumulated in big cities at higher rates than in rural areas, and consequently, surface temperature is higher in cities. This effect is best appreciated at nighttime due to the absence of direct heat due to solar radiation.
A special case of UHI is Extreme UHI (E-UHI), also known as heat point, which is located in industrial areas with extremely high temperatures [1]. These industrial areas are characterized by a lack of vegetation and the presence of big factory sheds, frequently hosting activities requiring high temperatures, such as steel plants and thermal energy installations. In industrial buildings, the parameter that determines the temperature rise is the material of the building envelope, mostly the material of the building roof, as this is the main area of heat emission from the inside but also of heat absorption from the outside.
Albedo is the fraction of incoming radiation that the surface reflects, with values from 0 to 1 for lowest and highest reflection, respectively. Albedo, the same as Land Surface Temperature (LST), is a key parameter for the study of the effect of UH and can serve as a measure for its mitigation. An example is the "Cool Roofs" [2], which are characterized by high values of albedo for high solar reflectance.
In addition, both parameters of albedo and LST can be studied on a large scale by using satellite imagery in the infrared spectrum, visible band and short-wave infrared for the albedo (0.2-3 µm), and thermal infrared for LST (7-14 µm).
This work studies E-UHI in industrial areas in Valladolid (Spain) as a case study for the determination of the correlation between albedo and LST, showing that the building roofs with high albedo values reach lower temperatures, as in [3], but by using satellite imagery instead of a spectrophotometer towards the design of a generalized methodology. the determination of the correlation between albedo and LST, showing that the building roofs with high albedo values reach lower temperatures, as in [3], but by using satellite imagery instead of a spectrophotometer towards the design of a generalized methodology.

Case Study
The city of Valladolid (Spain) was selected as representative for the study. There are 6 industrial areas in different locations within the city. Seven factory sheds were selected for the study, reaching a total surface of 40 ha ( Figure 1).

Materials
Albedo and LST were calculated using data acquired by satellite platforms. Particularly, images from Landsat-8 satellite (OLI and TIRS sensors) and Terra satellite (MODIS and ASTER sensors) were used with spatial resolutions of 30 m (Landsat8), 1000 m (MODIS), and 90 m (ASTER). OLI, TIRS, and MODIS were used to calculate daytime LST, ASTER provided a pre-processed image of nighttime LST, and OLI was used to calculate albedo. The reason for the use of different sensors in daytime/nighttime studies is the passing time of each platform: Landsat 8 passes over Valladolid at 11:00, meanwhile TERRA passes at 12:00 during daytime and also at 22:20 during nighttime. Further information about these sensors and their technical characteristics can be found in [4][5][6].

LST
LST was computed with the methodology presented in [7], which is based on single-channel methodologies ( Figure 2).

Materials
Albedo and LST were calculated using data acquired by satellite platforms. Particularly, images from Landsat-8 satellite (OLI and TIRS sensors) and Terra satellite (MODIS and ASTER sensors) were used with spatial resolutions of 30 m (Landsat8), 1000 m (MODIS), and 90 m (ASTER). OLI, TIRS, and MODIS were used to calculate daytime LST, ASTER provided a pre-processed image of nighttime LST, and OLI was used to calculate albedo. The reason for the use of different sensors in daytime/nighttime studies is the passing time of each platform: Landsat 8 passes over Valladolid at 11:00, meanwhile TERRA passes at 12:00 during daytime and also at 22:20 during nighttime. Further information about these sensors and their technical characteristics can be found in [4][5][6].

LST
LST was computed with the methodology presented in [7], which is based on singlechannel methodologies (Figure 2). the determination of the correlation between albedo and LST, showing that the building roofs with high albedo values reach lower temperatures, as in [3], but by using satellite imagery instead of a spectrophotometer towards the design of a generalized methodology.

Case Study
The city of Valladolid (Spain) was selected as representative for the study. There are 6 industrial areas in different locations within the city. Seven factory sheds were selected for the study, reaching a total surface of 40 ha (Figure 1).

Materials
Albedo and LST were calculated using data acquired by satellite platforms. Particularly, images from Landsat-8 satellite (OLI and TIRS sensors) and Terra satellite (MODIS and ASTER sensors) were used with spatial resolutions of 30 m (Landsat8), 1000 m (MODIS), and 90 m (ASTER). OLI, TIRS, and MODIS were used to calculate daytime LST, ASTER provided a pre-processed image of nighttime LST, and OLI was used to calculate albedo. The reason for the use of different sensors in daytime/nighttime studies is the passing time of each platform: Landsat 8 passes over Valladolid at 11:00, meanwhile TERRA passes at 12:00 during daytime and also at 22:20 during nighttime. Further information about these sensors and their technical characteristics can be found in [4][5][6].

LST
LST was computed with the methodology presented in [7], which is based on single-channel methodologies ( Figure 2).

Albedo
Albedo is a parameter with a complex computation due to the high number of variables of influence [8]. Thus, there are numerous methodologies and strategies for albedo computation. In this study, the methodology from Baldinelli was applied [9] because it was performed in an urban area, particularly for factory sheds (Figure 3). Baldinelli's methodol-Eng. Proc. 2021, 8, 9 3 of 4 ogy provides good results in urban areas because it takes into account the anisotropy of the surface. Albedo was computed in this case from Landsat8 imagery (OLI sensor).

Albedo
Albedo is a parameter with a complex computation due to the high number of variables of influence [8]. Thus, there are numerous methodologies and strategies for albedo computation. In this study, the methodology from Baldinelli was applied [9] because it was performed in an urban area, particularly for factory sheds (Figure 3). Baldinelli's methodology provides good results in urban areas because it takes into account the anisotropy of the surface. Albedo was computed in this case from Landsat8 imagery (OLI sensor).

Results
The results of the study show the existence of a negative correlation between albedo and daytime LST (Figure 4), with R 2 equal to −0.6109. This correlation confirms the theory that high values of albedo (reflected radiation) correspond to lower values of daytime LST (thus, lower heat absorption). A detailed view of the roofs under study shows that roofs with high albedo values and low daytime LST also present lower nighttime LST than roofs with low albedo (Figure 5). Color correspondence can also be seen, where dark roofs usually correspond to low albedo values and light roofs present high albedo. However, nighttime LST shows results independent from the value of albedo, mainly due to the thermal properties of the building interior and its capacity to retain heat.

Results
The results of the study show the existence of a negative correlation between albedo and daytime LST (Figure 4), with R 2 equal to −0.6109. This correlation confirms the theory that high values of albedo (reflected radiation) correspond to lower values of daytime LST (thus, lower heat absorption).

Albedo
Albedo is a parameter with a complex computation due to the high number of variables of influence [8]. Thus, there are numerous methodologies and strategies for albedo computation. In this study, the methodology from Baldinelli was applied [9] because it was performed in an urban area, particularly for factory sheds (Figure 3). Baldinelli's methodology provides good results in urban areas because it takes into account the anisotropy of the surface. Albedo was computed in this case from Landsat8 imagery (OLI sensor).

Results
The results of the study show the existence of a negative correlation between albedo and daytime LST (Figure 4), with R 2 equal to −0.6109. This correlation confirms the theory that high values of albedo (reflected radiation) correspond to lower values of daytime LST (thus, lower heat absorption). A detailed view of the roofs under study shows that roofs with high albedo values and low daytime LST also present lower nighttime LST than roofs with low albedo (Figure 5). Color correspondence can also be seen, where dark roofs usually correspond to low albedo values and light roofs present high albedo. However, nighttime LST shows results independent from the value of albedo, mainly due to the thermal properties of the building interior and its capacity to retain heat.  A detailed view of the roofs under study shows that roofs with high albedo values and low daytime LST also present lower nighttime LST than roofs with low albedo ( Figure 5). Color correspondence can also be seen, where dark roofs usually correspond to low albedo values and light roofs present high albedo. However, nighttime LST shows results independent from the value of albedo, mainly due to the thermal properties of the building interior and its capacity to retain heat.

Albedo
Albedo is a parameter with a complex computation due to the high number of variables of influence [8]. Thus, there are numerous methodologies and strategies for albedo computation. In this study, the methodology from Baldinelli was applied [9] because it was performed in an urban area, particularly for factory sheds (Figure 3). Baldinelli's methodology provides good results in urban areas because it takes into account the anisotropy of the surface. Albedo was computed in this case from Landsat8 imagery (OLI sensor).

Results
The results of the study show the existence of a negative correlation between albedo and daytime LST (Figure 4), with R 2 equal to −0.6109. This correlation confirms the theory that high values of albedo (reflected radiation) correspond to lower values of daytime LST (thus, lower heat absorption). A detailed view of the roofs under study shows that roofs with high albedo values and low daytime LST also present lower nighttime LST than roofs with low albedo (Figure 5). Color correspondence can also be seen, where dark roofs usually correspond to low albedo values and light roofs present high albedo. However, nighttime LST shows results independent from the value of albedo, mainly due to the thermal properties of the building interior and its capacity to retain heat.

Conclusions
The analysis of albedo and daytime LST in the roofs of factory sheds in the city of Valladolid shows a negative correlation between the two parameters in such a way that low albedo (dark color) roofs present higher LST than high albedo (light color) roofs. In addition, low albedo roofs also present higher nighttime LST.