Overview

Project Highlights:

 

  • Merging of LST data from different sensors to produce global temperature fields under all-sky conditions (cloudy and non-cloudy).
  • Understanding of differences between infrared and microwave surface temperature data and building of robust relationships.
  • Evaluation of other LST analyses and models with the new dataset.

 

Land surface temperature (LST) is a fundamental, spatial quantity which is a fascinating, emerging observable for environmental science. LST is a key boundary condition state variable in land surface models, which determine the surface -to- atmosphere fluxes of heat, water and carbon compounds and represents the boundary condition in climate models. It also influences cloud cover, precipitation and atmospheric chemistry predictions within these models. Model deficiencies in representing LST often provide an indication of problems in surface energy fluxes and soil moisture that can affect the actual performance of Earth System Models at various temporal scales. There is an increasing focus on the opportunity to exploit satellite LST data to confront the challenges of climate science; and our research group is leading the international effort on LST science.

LST can be determined from thermal emission in either infrared (IR) or microwave (MW) atmospheric windows. Infrared skin temperature is defined as the temperature measured by an IR radiometer in cloud-free conditions, typically operating at wavelengths 3.7-12 µm. It is the temperature of the top few micrometres of the surface (whether bare ground or canopy leaves), Microwave skin temperature represents the surface temperature at depths up to a few millimetres, depending on wavelength, view angle and surface conditions.

Retrievals in the IR are generally more accurate than MW retrievals due to smaller variation of surface emissivities and stronger dependence of the radiance on temperature. Nevertheless, microwave measurements have been shown to complement those in the IR due to their lower sensitivity to clouds, thus increasing sampling in cloudy conditions; although their spatial resolutions are significantly lower than their IR counterparts.

Use of LST for climate studies has been hindered because longer-term datasets are based on IR observations, which are limited to clear-sky. This presents a problem for many applications as resulting trends may be clear-sky biased, and climate models include both cloudy and clear-sky simulations. This project will progress our ability to overcome this by better understanding the physical differences between the observations and how robust relationships can be developed to enable observations to be merged into a consistent record.

 

 

Figure 1: Pre-cursor attempt to blend LST data from different polar-orbiting satellite sensors (see www.globtemperature.info)

Methodology

To meet the most significant LST requirements for climate science, integrated products can take advantage of the strengths of each data stream (IR and MW; polar orbiting and geostationary; and where available, in situ). The ultimate aim is to provide sub-daily, near-global coverage to better understand the diurnal (24hr) variability in LST. The first steps are to perform some initial comparisons of the IR and MW datasets and then to develop a draft process for merging the data. Much of the detailed research will develop from these key factors and will establish robust relationships between temperature measured from the two different techniques as a function of other parameters such as cloud thickness and wind speed. The results will improve IR and MW retrievals, build a definitive LST dataset derived from the individual data sets and utilise the product to evaluate climate models and other surface temperature datasets.

Training and Skills

In the first year, students will be trained on environmental data science, research methods and core skills. Throughout the PhD, training will progress from core skills sets to master classes specific to this project’s themes. Specialist training will include sensor techniques, radiative transfer for infra-red and microwave, non-linear data methods and general remote sensing. The National Centre for Earth Observation will provide access to its Researcher Forum, staff conferences/workshops and national-level training. There is good access to international summer schools.

Timeline

Year 1: Training in software usage and development, and attendance at dedicated workshops. Initial evaluation of IR and MW data and data analysis to produce a first merged dataset.

Year 2: Determination of robust relationships and construction of blending approach. Conference attendance, preparation of manuscript for journal submission. Continued development of thesis chapters.

Year 3: Merged product suitable for climate studies and evaluation of models.. Manuscript submission and revision, International conference attendance, thesis preparation.

Partners and collaboration (including CASE)

This project will fall within the National Centre of Earth Observation (NCEO), which is the leading collective of satellite remote sensing in the UK. Furthermore, the project will feed directly into the framework of the Climate Change Initiative programme run by the European Space Agency (ESA). The student will have an excellent opportunity to work alongside the leading scientists across Europe and beyond in measuring the temperature of the Earth from space.

Further Details

Professor Remedios is Director of NCEO and Professor at University of Leicester. NCEO is a national centre funded by NERC and distributed across key Earth Observation (EO) groups at Universities and research laboratories. Dr Darren Ghent is the lead scientist on the international Climate Change Initiative LST Project, and leads the LST activities for the operational Sentinel-3 satellite mission. The student will have a chance to be part of a national EO community complementing the environmental science focus of CENTA.