Fast Retrieval of Land Surface Emissivity from Landsat Data through IDL Programming

Land surface emissivity (LSE) has been an essential parameter for many quantitative thermal infrared remote sensing models. There are three main ways to acquire it: laboratory measurement, field survey and remote sensing retrieval. Due to the inconvenience of in-situ measurements as well as the limited surface coverage of the former two methods, remote sensing retrieval has become the major option for various studies with its incomparable advantages of large instantaneous spatial coverage and repeated reliable measurements. As of today, a lot of algorithms for LSE retrieving have been proposed. However, no matter what kind of algorithm is used, the process of LSE retrieving is always quite fussy and time-consuming, because there are many other involving parameters (such as: LULC types, land surface reflectance and vegetation fraction, etc.) need to be acquired or retrieved in advance. Currently, the most often used way to retrieve LSE is to utilize remote sensing software like ENVI (or ERDAS Imagine, PCI Geomatica, IDRISI, etc.) step by step with a very long, exhausting process of data selection and equations inputting, which made people very prone to make mistakes. The aim of this work was to find an easy-to-use, quick and efficient way to simplify the process of LSE retrieval. For that purpose, the IDL (Interactive Data Language) programming was used and the LSE retrieving algorithm proposed by Qin (2004) was introduced and applied. By using the plug-in for ENVI (in.sav format), the whole procedure became very simple, only input and output selections were needed in the user-interface. The complex computing and selection jobs were sealed and left no chances for mistakes like typing errors when inputting complicated equations. A subset of a Landsat ETM+ imagery (path 119, row 42, acquired on May 29, 2003) was used to test the plug-in and a satisfying result was achieved, the LSE was calculated within seconds. In the final session of the present paper, a discussion on how to modify the corresponding codes to make the plug-in suitable for other kinds of remote sensing data (e.g., Landsat 8) was made.