Projection of land surface temperature considering the effects of future land change in the Taihu Lake Basin of China

Abstract

Land surface temperature (LST) is an important environmental parameter that is significantly affected by land use and landscape composition. Despite the recent progress in LST retrieval algorithms and better knowledge of the relationship between LST and land coverage indices, predictive studies of future LST patterns are limited. Here, we project LST patterns in the Taihu Lake Basin to the year 2026 based on projected land use pattern and simulated land coverage indices that include normalized difference built-up index (NDBI), normalized difference vegetation index (NDVI) and normalized difference water index (NDWI). We derived the spatiotemporal LST patterns in the Taihu Lake Basin from 1996 to 2026 using thermal infrared data from Landsat imagery. A CA-Markov model was applied to project the 2026 land use pattern in the basin based on spatial driving factors, using the 2004 land use as the initial state. We simulated the NDBI, NDVI and NDWI indices for 2026 using the projected land use patterns, and then generated the 2026 LST in the study area. Our results showed that LST has been increasing and the warming areas have been expanding since 1996, especially in the Su-Xi-Chang urban agglomeration. The mean LST in Su-Xi-Chang has increased from less than 30°C in 1996 to greater than 31°C in 2004 and has risen to about 33°C in 2016, and the projection suggests that LST will reach about 35°C in 2026. Our results also suggest that mean LST increased by 2°C per decade in this highly urbanized area between 1996 and 2026. We present a preliminary method to produce future LST patterns and provide reasonable LST scenarios in the Taihu Lake Basin, which should help develop and implement management strategies for mitigating the effects of urban heat island.

Spatial Patterns of Land Surface Temperature and Their Influencing Factors: A Case Study in Suzhou, China

“…China has experienced rapid urban growth, with extensive expansion of high-density built-up areas [38,49] and large buildings with reinforced-concrete floors, in the city center. With increased demands for improved living space quality and the reconstruction of the downtown areas, more green space has been built in the city center, recently, slightly alleviating the SUHI effects in downtown Suzhou.…”

“…These reflective bands were used to produce land coverage indices and land-use patterns. LST at Suzhou was derived using the Landsat-5 TM thermal band-6 and Landsat-8 TIRS band-10, based on the radiative transfer equation, as described in early publications [38,51]. The thermal infrared radiation ( ) can be given by [38,51]:…”

“…We calculated vegetation coverage and land surface emissivity for atmospheric radiation correction [63,64], and converted surface heat radiation intensity to LST [65,66]. We derived the spring LST data (Figure 8a-c) using the method presented in subsection 2.2 and used the summer LST data (Figure 8d-f), reported by an earlier publication [38] on the Taihu Lake Basin. The LST maps clearly show that the warmest regions lie in and near the city center and have expanded in spatial extent, with urban growth (Figure 8).…”

“…Modelers have applied the three indices and quantitatively described the impacts on LST, which is recognized to be positively correlated with NDBI but negatively correlated with NDVI and NDWI [36,37]. The three indices have also been applied in predicting future LST patterns, by establishing their quantitative relationships [38].Landscape patterns and composition have been found to have close relationships with LST, and with SUHIs [39]. Landscape composition is a key factor in distinguishing the effects of land class Remote Sens.…”

“…Modelers have applied the three indices and quantitatively described the impacts on LST, which is recognized to be positively correlated with NDBI but negatively correlated with NDVI and NDWI [36,37]. The three indices have also been applied in predicting future LST patterns, by establishing their quantitative relationships [38].…”