Thermal Directionality Study and Application of Thermal Radiation to Drought Monitoring

Abstract

Remote sensing (RS), with its large spatial coverage and easily accessible observations, has attracted a lot of attention in recent years. Thermal infrared (TIR) RS, collecting radiation between 3.75 and 12.5 μm in the electromagnetic spectrum, is one of the major parts of RS. TIR RS is widely used in various fields, including evapotranspiration (ET), global climate change, hydrological cycle, vegetation monitoring and urban climate given the important role TIR radiation plays in surface energy and water balance. TIR radiation is closely related to land surface temperature (LST) and land surface emissivity (LSE). Angular variation is an important characteristic of LSE, which could influence the subsequent estimation of surface upwelling longwave radiation (SULR) and LST. In this study, a look-up table (LUT) of directional emissivities was built from the MYD21A product. The compiled LUT was then applied to SULR and LST estimation by considering the angular variation of LSE. The results showed that the influence of LSE angular variation on SULR estimation was not pronounced. Whereas, the influence on LST retrievals was > 0.5 K and the accuracy of the split-window (SW) was improved by > 1 K over barren surfaces after considering LSE directionality. LST is connected to ET through the surface energy balance equation, thereby reflecting vegetation water availability. In this study, applying TIR radiation in agricultural drought early warning was of interest. Based on the underlying principle that the rate of LST rise between 1.5 and 3.5 h after the sunrise is approximately linear and over vegetated surfaces occurs more rapidly under dry conditions as a consequence of stomatal control, the temperature rise index (TRI) was developed using the LST retrievals from the geostationary Multifunction Transport Satellite-2 (MTSAT-2) instrument and using the Himawairi-8 brightness temperatures (BT), respectively. The proposed TRI was evaluated by comparing with more commonly-used indices, including precipitation condition index (PCI), soil moisture condition index (SMCI) and vegetation condition index (VCI). In addition, the indices were also compared to annual wheat yield over large areas in the Australian Wheatbelt. The results showed that the TRI produced spatiotemporal dryness patterns that were very similar to those in soil moisture and precipitation, but with more detail due to its finer resolution. A time lag was found between TRI and observed vegetation condition, supporting the use of TRI in early warning. Among the compared drought indices, the TRI had the strongest and earliest correlation with wheat yield. The TRI calculated from LST and BT had close performances. It is concluded that this study provides insights into the basic theory study as well as practical applications of TIR RS, and adds value to the state-of-the-art studies in the field of TIR RS.