Modeling the Impact of Surface Characteristics on the Near Surface Temperature Lapse Rate

Abstract

Abstract. Modeling of Near-Surface Temperature Lapse Rate (NSTLR) is very important in various environmental applications. The Land Surface Temperature (LST) is influenced by many properties and conditions including surface biophysical and topographic characteristics. Some researches have considered the LST - Digital Elevation Model (DEM) feature space to model NSTLR. However, the influence of detailed surface characteristics is rare. This study investigated the impact of surface characteristics on the LST-DEM feature space for NSTLR modeling. A set of remote sensing data including Landsat 8 images, MODIS products, and surface features including DEM and land use of the Balikhli-Chay on 01/07/2018, 18/08/2018 and 03/09/2018 were collected and used in this study. First, Split Window (SW) algorithm was used to estimate LST, and spectral indices were employed to model surface biophysical characteristics. Owing to the impact of surface biophysical and topographic characteristics on the LST-DEM feature space, the NSTLR was calculated for different classes of surface biophysical characteristics, land use, and solar local incident angle. The modeled NSTLR values based on the LST-DEM feature space on 01/07/2018, 18/08/2018 and 03/09/2018 were 8.5, 1.5 and 2.4 °C Km−1; respectively. The NSTLR in different classes of surface biophysical characteristics, land use type and topographical parameters were variable between 0.5 to 14 °C Km−1. This clearly showed the dependence of NSTLR on topographic and biophysical conditions. This provides a new way of calculating surface characteristic specific NSTLR.

Analysis of Near-Surface Temperature Lapse Rates in Mountain Ecosystems of Northern Mexico Using Landsat-8 Satellite Images and ECOSTRESS

“…The high variation observed in summer NSTLR across ESI intervals, with a range from wetter areas with lower NSTLR, to areas with higher water stress and with higher NSTLR values, might be related to contrasting interactions of vegetation cover and water stress conditions. Areas with high vegetation cover limit sunlight incidence on terrain surface, cooling surfaces through evapotranspiration, while bare ground areas are associated to high surface temperatures; this might result in high NSTLR, mainly in summer season, where the solar radiation has higher effects [18,26]. While our findings agree with previous studies reporting effects of water availability on LST in environments with altitudinal variability (e.g., [1,27]), such previous analyses were conducted utilizing other water availability variables such as NDVI [1], unlike the ECOSTRESS ESI data utilized here.…”

“…NSTLR can be affected by topographic variables, such as aspect and slope and other variables such as evapotranspiration (ET) [1,25], although many questions remain regarding the environmental drivers of NSTLR. The characterization of NSTLR in mountain ecosystems is limited by the availability and spatial distribution of weather stations, often limiting a spatially detailed characterization of climatic data [24,26,27].…”

“…Several studies have been utilized satellite information to estimate and analyze the NSTLR in mountain ecosystems, using the relationship between LST and a digital elevation model (DEM) [8,18,26,[31][32][33]. Nevertheless, it has been observed that some environmental variables (i.e., solar incidence, moisture, land cover characteristics) cause interference of NSTLR estimated from LST data [18,34].…”

“…This approach has been less explored and requires further evaluation, particularly for the relatively less utilized medium resolution sensors. In particular, studies using Landsat Land Surface Temperature product to analyse factors driving thermal gradients in forest ecosystems are still relatively scarce [18,26,36]. In the present study, we used LST derived from Landsat-8 TIRS, which was normalized using random forests to obtain NSTLR information.…”

“…In addition to the previously documented contrasting effects of vegetation and topography on NSTLR [18,26], other parameters potentially affecting thermal gradients have received less attention. For example, very few studies have analysed the variation of ther-mal gradients with respect to remotely sensed evapotranspiration (ET) indices, particularly from medium resolution sensors.…”