At any point in time, hundreds of satellites are moving around the earth, imaging it at different locations and intervals. The satellites receive electromagnetic waves which are reflected from the earth surface creating a signal. Depending on the nature of the surface, incident waves are reflected in different ways; analysing these reflectance patterns can reveal numerous patterns across space and time.
The science of remote sensing holds immense potential when we wish to look at large scale patterns which are too difficult to be covered on the ground. One such application is measuring “land surface temperature” (LST). LST is different from the ambient temperature that we usually come across in weather reports; it is the temperature of the top layer of the earth’s surface and has a crucial influence on ecology, hydrology, agriculture and even the ambient temperature. Through remote sensing, LST has been recorded for a long time now. “Usually these measurements are retrieved from thermal remote sensors, but these sensors fail to provide LST data under cloudy conditions”, says the author.
Infrared radiation in the thermal remote sensors gets absorbed by clouds & water vapour and therefore there is no signal going back to the satellite, leading to missing information. With global warming becoming an increasingly serious problem, scientists have been trying to find ways to fill these data lacunae. In such scenario, microwave sensors can complement the available thermal sensors. The advantage of using microwaves over infrared is that they can penetrate through non-precipitating clouds, thus eliminating the possibility of gaps in LST data when using infrared.
This method for land surface temperature prediction has been applied by scientists at the Indian Institute of Science, Bangalore, in their recent research exploring the use of microwave remote sensing in addition to traditional thermal remote sensing to predict LST over the Cauvery river basin in South India.
The predictions were made using artificial neural networks (ANN), a statistical algorithm based on biological nervous systems capable of learning and pattern recognition. The microwave satellite data was obtained from advanced microwave scanning radiometer (AMSR-E) on-board Aqua satellite.
Microwave vegetation indices improve the results of LST prediction because they are sensitive to vegetation parameters, with less interference from soil and atmospheric components as compared to traditional vegetation indices. The researchers validated the potential of microwave based LST prediction by comparing it with LST values measured when there were no clouds, by the well-established infrared based moderate resolution imaging spectroradiometer (MODIS) sensor; the LSTs obtained were indeed comparable.
The major disadvantage of microwave based LST estimation however, is the coarse spatial resolution of microwave data, which can influence the precision of the LST values. Nevertheless, it is a major advancement in the field of remote sensing science with high applicability in areas where cloud conditions obstruct traditional LST measurements.
About the authors
Prof D Nagesh Kumar and Shwetha H R are at the Department of Civil Engineering, Indian Institute of Science, Bangalore.
Tel: 080-2293-2666; E-mail: firstname.lastname@example.org
About the paper
The work was presented at the International Conference on Water Resources, Coastal and Ocean Engineering (ICWRCOE 2015) at NIT, Suratkal earlier this year and it appeared as a proceedings paper in Aquatic Procedia (10.1016/j.aqpro.2015.02.179).