| dc.description.abstract | Surface energy processes have an essential role in the regional weather, climate, and hydrological cycles. Understanding the characteristic of energy exchange processes is vital to improve the efficiency and effectiveness of many applications, such as water resources management, land use planning, forest vulnerability assessment, numerical modeling of hydrological processes, weather forecasting system, etc. Ground-based instrument systems have been long used as a point-wise measurement method for land heat fluxes estimation. However, their use is often restricted to a pre-defined domain and limited with spatial coverage, resulting in difficulty to operate over a regional scale. With the development of satellite-based systems, space science and technology is a key component of the global observation system, which provides a promising data source for the investigations of land-air heat exchange with high spatio-temporal efficiency and economic benefits. Different methods have been proposed and performed with varying degrees of input requirement from remotely sensed data in which the approaches of physical processes in surface energy balance models were widely used to estimate the energy exchange at the land-air interface using the remote sensing data as the primary input. However, these approaches involve the establishment of complex physical processes and still are limited to the areas with existing ground-based measurement system.
Multi-spectral remote sensing has been widely utilized to collect the Earth’s surface properties that basically influence the processes of heat transfer at the land-air interface. The distinction in reflectance makes it possible to understand the earth′s surface features by analyzing its spectral reflectance signatures. In this study, a new multiple-band index, Normalized Difference Latent heat Index, is proposed for latent heat flux extraction from satellite imagery. It utilizes the reflectance observations of three channels, which are primarily used for the optical Earth observation satellites, including green, red, and shortwave-infrared. The NDLI is firstly applied to extract the information of land–air latent heat exchange over a subset region in the eastern part of Chiayi City, Taiwan. The NDLI exhibited superiority in representing the potential latent heat flux by showing the highest consistency coefficient (r = 0.75) with the corresponding predicted results from the Surface Energy Balance Algorithm for Land model (SEBAL) as compared to the other existing satellite indices. Subsequently, the NDLI is adopted to enhance the evapotranspiration estimation over rice paddy areas in the Thai Binh Province, Vietnam. Results indicated that the NDLI-derived ET differs from the derivative of SEBAL by less than 10% over 98.1% of the paddy field. Moreover, the NDLI-derived ET exhibited its superiority in revealing the low amounts of ET in the rice paddy areas under stress. It is concluded that the newly developed NDLI is a good indicator to quantify the surface moisture, evapotranspiration, and latent heat transfer at the land-air interface. The NDLI is potentially helpful for a variety of practical applications since its formula is simple and easily implemented from the abundant existing satellites | en_US |