You are here

Scientists design new methodology to better predict Land Surface Temperature under cloudy conditions

Measurement of Land Surface Temperature (LST) is important for hydrology, environment science and many associated fields. Prof. D. Nagesh Kumar of Department of Civil Engineering at the Indian Institute of Science, Bangalore, and his team have proposed a novel methodology that can accurately predict LST at high resolution even under cloudy conditions. Their work has been published in the ISPRS Journal of Photogrammetry and Remote Sensing.

Land Surface Temperature is the radiative skin temperature of the uppermost part of the earth's surface. Its value may differ from the ambient air temperature. "Land Surface Temperature and its difference with the air temperature is an important parameter that is needed for various applications that include evapotranspiration estimation, climate change, flood and drought prediction and environmental studies. LST varies spatially and temporally. If we have temperature measurements with fine resolution, we can make better predictions about rainfall and floods", explains H.R. Shwetha, one of the team members.

The researchers used sensor data from Microwave Antennae installed on satellites. Though installing sensors on the land is the most accurate method to measure temperature, it is impractical to provide values over wide areas. Remote sensing from satellites provides a good way to obtain high resolution temperature measurement. Infrared satellite sensors can be used to sense temperature, but these sensors do not work under cloudy conditions. Microwaves can penetrate through clouds, facilitating measurement in cloudy conditions. Though remote-sensing using microwaves is not new, it was never before used in India for LST estimation. 

The researchers chose the Cauvery River basin that occupies portions of Karnataka, Kerela, Tamil Nadu and Pondicherry, for the study. They used data from NASA's polar synchronous satellite Aqua. It carries sensors called MODIS (Moderate Resolution Imaging Spectraradiometer) and AMSR-E (Advanced Microwave Scanning Radiometer). They resampled readings for 1km resolution. Other data like latitude, longitude and altitude is also input for the sample points. They employed Artificial Neural Network (ANN) based models for different land cover classes and obtained relations between various parameters under clear sky conditions. With the assumption that these relations will hold good during cloudy conditions, they could then predict land surface temperatures under cloudy conditions.

The researchers found that the land surface temperature values predicted by the proposed method correlated well with the actual measured surface temperature values. Certain parameters of the evaluation process change with the land type, i.e. whether the land is arid or vegetated. Cauvery basin mostly has forests and crop-lands, so the proposed method is tested for land that has some vegetation.

"The proposed methodology is the most feasible way to predict LST at high spatio-temporal resolution under cloudy conditions in the absence of in-situ LST measurements at all land cover classes during daytime and night time," concludes Prof. Nagesh Kumar.

The team is working on facilitating the usage of the predicted LST in the estimation of evapotranspiration and soil moisture over the study region. This is an important phenomenon in the water cycle. Knowledge about this will help better rainfall prediction.

 

About the authors

Prof. D Nagesh Kumar is Professor in Dept of Civil Engg and also Chairman, Center for Earth Sciences, Indian Institute of Science, Bangalore

Shwetha H.R. is a PhD student in the Dept of Civil Engineering, Indian Institute of Science, Bangalore

 About the publication

Prof. Nagesh Kumar and Shwetha H.R. co-authored a paper named "Prediction of high spatio-temporal resolution land surface temperature under cloudy conditions using microwave vegetation index and ANN". It appeared in the journal "ISPRS Journal of Photogrammetry and Remote Sensing", DOI: 10.1016/j.isprsjprs.2016.03.011