Cloud and Cloud Shadow Detection and Removal Using Landsat-8 Data and Generation of Cloud-Free Image in Beas River Basin of Himachal Pradesh

Six successful Landsat missions have launched since 1972, contributing to nearly 50 years of earth observation to track the dynamics of land cover and change. On February 11, 2013, the Landsat Data Continuity Mission (LDCM), also known as Landsat-8, successfully launched, fulfilling its goal of collecting Earth imagery under an open data policy. Landsat-8 has two push-broom sensors. The first instrument, the Operational Land Imager (OLI), covers 185 km with 30 m of spatial resolution across nine shortwave spectral bands except for a 15-m panchromatic band. Two thermal bands—band 10 and band 11—were imaged by the Thermal Infrared Sensor (TIRS) spanning 185 km with a 100-m resolution. However, clouds and their shadows make it hard to see the ground surface and limit change pattern investigation. This study shows a new way to detect and remove clouds and cloud shadows using Landsat-8 image data (WRS2: Path/Row = 147/38, obtained on April 24, 2015). This approach employs six of the nine bands for alteration to compute the strength of clouds and cloud shadows, which are then eliminated. The spectral data from the different bands is used in this way. The validation indicates that the cloud shadow and polluted cloud pixels are accurately detected, achieving overall accuracies of 97 and 96%, respectively. Nonetheless, the efficacy of this approach is limited in the presence of thick clouds and cloud shadows. This technique should be improved to account for the weather and make it simpler to identify changes in the terrain.