Modelling nighttime air temperature from remote sensing imagery and GIS data

Abstract

The estimation of the Earth's surface temperature (LST) from the infrared thermal radiation (TIR) emitted by the Earth detected by means of remote sensing has allowed a giant leap in climate analysis. The specialized literature has highlighted the singular importance of the LST in the generation of the Urban Heat Island (UHI), especially at night. It is during the night that the effects of UHI become more apparent, due to the low cooling capacity of urban construction materials and is during nighttime that temperatures can cause higher health risks, leading to the aggravation of negative impacts on people’s health and comfort in extreme events such as heat waves becoming more and more frequent and lasting longer. However, the study of nocturnal UHIs is still poorly developed, due to the structural problems. On the one hand, the scarcity of meteorological stations that allow obtaining the air temperature (Ta) with an adequate degree of spatial resolution. And, on the other, due to the limited temporal availability of medium / high resolution nocturnal satellite images that make it possible to know the LST at night. Traditional methods for obtaining nocturnal UHI have been directed either to extrapolation of data from weather stations, or obtaining Ta through urban transects. In the first case, the lack of weather stations in urban landscapes makes it extremely difficult to obtain data to extrapolate and propose models at a detailed resolution scale. In the second case, there is a manifest difficulty in obtaining data simultaneously and significantly representative of urban and rural zones. Remote sensing images are another methodology used to measure nighttime UHI, but the greatest limitation of this method is the scarcity of high-resolution images that allow rigorous nighttime LST to be obtained. Only MODIS, or Sentinel 3, offer free mid-resolution nighttime thermal imaging for LST and UHI analysis. Furthermore, the integration of LST (obtained from remote sensing imagery) with Ta (obtained from weather stations) continues to be a pending challenge. The right estimation of the temperature of the air at ≈ 2-m height above ground (Ta) from LST is possible but complex. The vertical lapse rate to be applied is function of the surface energy balance, which varies in function of the nature of the surface and of the instant of the day. The objective of this paper is to integrate the information derived from the thermal band of satellite images (LST) with the "in situ" measurements of the Ta obtained at meteorological stations. The methodology used consists, first, in developing a model by means of multi-regression analysis of the night air temperatures, using as explanatory variables, in addition to the physical characteristics of the territory (longitude, latitude, altitude, distance to the sea, slope, orientation), the characteristics derived from vegetation (NDVI), urbanization (NDBI) as well as the LST obtained by means of MODIS. And, secondly, downscaling the previous model, using the information obtained from Landsat 8. In this way, a set of models is obtained at different resolutions that allow estimating the nighttime temperature at a detailed level. The case study is the Metropolitan Area of Barcelona (636 km², 3.3 million inhabitants).