TRISHNA: AN INDO-FRENCH SPACE MISSION TO STUDY THE THERMOGRAPHY OF THE EARTH AT FINE SPATIO-TEMPORAL RESOLUTION

J.-L. Roujean (1), B. Bhattacharya (2), P. Gamet (1), M.R. Pandya (2), G. Boulet (1), A. Olioso (3),
S.K. Singh(2), M. V. Shukla(2), M. Mishra(2), S. Babu(16), P. V. Raju(15), C.S. Murthy(15), X. Briottet (4),
A. Rodler (5), E. Autret (6), I. Dadou (7), D. Adlakha(2), M. Sarkar(2), G. Picard (8), A. Kouraev (7), C. Ferrari (9), M. Irvine (10), E. Delogu (11), T. Vidal (12), O. Hagolle (1), P. Maisongrande (11), M. Sekhar(14), K. Mallick(13)

(1) CESBIO, Toulouse, France (9) IPGP, Paris, France
(2) SAC, ISRO, Ahmedabad, India (10) INRAE, Bordeaux, France
(3) INRAE, Avignon, France (11) CNES, Toulouse, France
(4) ONERA, Toulouse, France (12) ACRI, Toulouse, France
(5) CEREMA, Nantes, France (13) LIST, Luxembourg
(6) LOPS, Plouzané, France (14) IISC, Bengaluru
(7) LEGOS, Toulouse, France (15) NRSC, ISRO, Hyderabad
(8) IGE, Grenoble, France (16) SPL, ISRO, Trivandrum


ABSTRACT

TRISHNA (Thermal infraRed Imaging Satellite for High-Resolution Natural resource Assessment) is a cross-purpose high spatial and temporal resolution thermal infrared Earth Observation (EO) mission that will provide observations in the domains of terrestrial and marine ecosystems, inland and coast, urban, cryosphere, solid Earth and atmosphere. It is an Indo-French innovative polar-orbiting mission that will overcome the limitations of TIR-optical observations from Landsat series and ASTER. The high-quality optical-thermal imagery will be used to provide precise surface temperature, emissivity and albedo of vegetation, manmade structures, snow, ice and sea. Atmospheric fields such as cloud mask and type, aerosol load, water vapour content will be well described. TRISHNA products will improve our knowledge of radiative and heat transfer to quantify evapotranspiration, fresh water discharge, snow-melt runoff, bio-geochemical cycle, urban heat island.
Energy transfer and exchanges of water and carbon fluxes in the soil–vegetation–atmosphere system need to be well described to enhance the role of environmental biophysics. Climate indicators include the surface temperature, the ocean heat, the glaciers and the Arctic and Antarctic sea ice extent. Land surface temperature (LST) and land surface emissivity (LSE) are Essential Climate Variables (ECV) (Global Climate Observing System/GCOS). LST is defined as the radiative skin temperature and is useful in agriculture (plant growth, water stress, crop yield, precision farming, early warning, freezing scenarios), hydrology (water cycle, catchment water, etc), and meteorology. LSE differentiates the surface attributes (vegetation, soil, snow, water, rock, manmade material) composing the landscape.
Water use in agriculture represents 70 % of global resources, making sustainable irrigation a key issue. Automatic detection and mapping of irrigated farmland area is vital for many services in charge of water management. In that respect, TIR signal brings, in addition to visible and near infrared, key information on irrigated areas that display lowest LST values at the peak of growth. The global change imposes an implementation of more efficient irrigation practices at the scale of an agricultural plot for better control. The decrease of moisture within the soil after water supply can be evaluated from the surface moisture estimated by radar but TIR observations remain better-suited to monitor vegetation water stress and irrigation at the agricultural plot to adapt the proper needs of each of the cultures. With a pixel size of 57 m, a revisit of 3 days at noon-night time, 6 VNIR (blue, green, red, NIR, water vapor, cirrus) and 4 TIR Bands in 8 – 12 m, TRISHNA will bring new insights. High resolution thermography is of broad interest for manifold domains like coastal area, inland water, urban areas, cryosphere, solid Earth and atmosphere. With a launch in 2025, TRISHNA will pioneer fine and routine collection of TIR scenes of the entire Earth owing to its instrumental design which proposes acquisitions based on across track scanner for 4 TIR channels (8,6, 9.0, 10.6 and 11.6 µm) and on push-broom.for 7 VNIR channels (485, 555, 670, 860, 910, 1380 and 1610 nm).
TRISHNA mission will look at early warnings of water scarcity and fire risk, including optimization for agricultural irrigation, rainfed crops, water consumption and its partitioning, plus food security. LST is a surrogate for soil water, near-ground air temperature and indirectly for productivity. TRISHNA will help to monitor any reduction and increase in evaporation (E) and transpiration (T). Evapotranspiration (ET) is an ECV that reflects the satisfaction or not of plant water needs. Its accuracy and timeliness is central as ET governs soil moisture at the surface and in the root zone through water extraction by plants, with large consequences for infiltration, runoff and for the whole catchment water budget. The detection of a water stress, deduced from a temporal chronicle, is useful to manage irrigation or to warn of the potential lack of water threatening ecosystems.
A main objective is to describe mixing processes and water quality in the coastal areas and estuarine from high-resolution SST (Sea Surface Temperature), also to enhance productivity and biodiversity assessment on the coast and for rivers and lakes including warning and monitoring of water borne diseases, to estimate energy fluxes in alluvial plains and aquifers, and to describe hazards (river floods, storm surges, inundated vegetation) in link to sea level rise.
Another main objective is to predict and to quantify at short-term the effects of the Urban Heat Island (UHI) on population health (Figure 5). A spectral un-mixing method was developed to provide sub-pixel abundances and temperatures from radiance images. Main goal is to improve LST retrieval due to enhanced footprint, to account for cavity anisotropy and environment effects, and to relate LST to air temperature. Urban areas are formed by complex 3D structures of mixed materials (cement, steel, bituminous roads, stones, bricks, glass, wood, grass, etc) and LST can vary locally by a few K. Topics are the modeling of urban climate, hydrology, building stress, storm water flow and generally UHI.
Key scientific question for the cryosphere concerns the combination of thermal and optical high resolution data to improve the prediction of the energy budget and the melting of snow and ice covered areas. In this regard, LST will be of added-value and how it may capture small-scale variability in mountainous areas is a relevant issue. Other issue will concern the mapping of debris-covered glaciers, lake ice formation and development, and lake water dynamics.
Snow is a good thermal insulator regulating soil temperature and sea-ice growth. Where the soil is permanently frozen (permafrost) presence of snow is critical for the preservation of the carbon storage in the soil. Besides, the evolution of the snowpack is a driver of the hydrological cycle in many watersheds.
Thermal anomalies may serve to anticipate volcano eruption and TIR measurements may help better characterizing volcanic ash clouds, geothermal resources, coal fires. Moreover, topography and roughness affect the surface energy balance of the solid Earth. Main properties are grain size, porosity, water content and composition. Soil temperature is a primary tracer of energy and water exchanges between the surface and the ground.
Downwelling shortwave and longwave radiation at noontime will be quantified under all-sky conditions through retrieval of aerosol optical depth, columnar precipitable water, cloud mask, cloud type / phase and albedo using seven optical and four TIR bands. In addition, surface air temperature and dust aerosol index using TIR data will be developed. These will be used for understanding aerosol-cloud interaction, to improve NWP model skills, agro-meteorological applications and air quality monitoring. Assimilation experiments will be carried out primarily with different land surface and atmospheric products as well as radiance assimilation to evaluate skill of NWP model weather forecast at various space-time scales.
CalVal is an important component of TRISHNA program. It consists in developing strategies to support the validation of TRISHNA Level 1 and 2 products, namely LST, LSE and ET. Efforts concern the data preprocessing to remove or minimize turbulent and directional effects that jeopardize the specification of 1 K on LST. Micrometeorological stations in different climates are equipped of TIR OPTRIS cameras and flown by UAV. Metrics and statistics are proposed as criteria for inter-comparison.
Cloud mask, atmospheric and anisotropy corrections (relief, directionality) are under development. Methods for measuring LST and LSE use consolidated approaches such as TES (Temperature-Emissivity Separation). Such key variables along with ET product will include accuracy assessment and a Quality Flag. ATBD are in preparation, notably the method of LST normalization. TRISHNA products will have global users such as FAO, GEOGLAM, global water watch, for meeting several Sustainable Development Goals (SDGs) as outlined by United Nations. Moreover, TRISHNA activities serve the preparation of future fine resolution TIR missions such as LSTM and SBG.

Evapotranspiration (ET) is a fundamental element of the hydrological cycle which plays a major role on both surface water and energy budgets. At local scale, ET can be estimated from detailed ground observations, for example using flux towers, but these measurements are only representative of a very limited number of homogeneous areas. When regional information is required, e.g. for monitoring ground water resources, ET can be mapped using thermal infrared and spectral reflectance data. Various ET models have been developed, but they often provide estimations within a large range of variations, suggesting high uncertainties in ET estimates.

We have developed the EVASPA (EVApotranspiration monitoring from SPAce) framework for estimating ET together with an estimation of its uncertainty. EVASPA is based on a multi-model multi-data ensemble system that provides maps of ET, global uncertainty and the contribution of each factors (models, input data) to the global uncertainty. EVASPA includes different procedures (or models) for estimating ET based on evaporative fraction formulations of the surface energy balance equations and/or based on aerodynamic equations. The system requires various data, such as surface temperature, solar radiation, air temperature, wind speed, surface albedo, leaf area index (LAI)…, as inputs. EVASPA considers several sources of data for each model input and the variability of input data is used to estimate input uncertainties. Overall, ET estimates and uncertainties are obtained by averaging and analyzing the variability of the different calculations of ET based on the different models and the different inputs.

In this study, EVASPA was applied to airborne data acquired over the Grosseto area in Italy in the frame of the ESA SurfSense experiment (high spatio-temporal Resolution Land Surface Temperature Experiment) in support of the LSTM mission project (Land Surface Temperature Monitoring). Surface temperature data were estimated using the Temperature Emissivity Separation (TES) method and considering different sources of atmospheric profiles of temperature and vapor. The TES method was applied on LSTM-like data which were constructed from the multispectral measurements performed with the TASI instruments. Meteorological data were obtained from different re-analysis products and ground measurements. Albedo and LAI were derived from reflectance measurements with the HyPlant instrument using different algorithms.

This analysis showed that the uncertainties on ET estimations might be very large up to 4 mm d-1 for the aerodynamic models and 3 mm d-1 for the evaporative fraction models. However, a better screening of inputs data and model formulation validity made it possible to decrease uncertainties down to 3 mm.d-1 (aerodynamic models) and 1.5 mm.d-1 (evaporative fraction models). Main uncertainty sources were related to solar radiation estimates, ground heat flux estimates, model formulations, in particular for the calculation of the evaporative fraction and the parameterization of the roughness length for heat transfer (aerodynamic models). In the case of aerodynamic models, wind speed and air temperature had also a significant impact, which explains the higher uncertainties for these models. In the frame of the development of future thermal infrared missions such as LSTM or TRISHNA, it is important to notice that the sensitivity of ET estimates to the uncertainty in the derivation of surface temperature from the thermal infrared data was lower than the impact of the other sources of uncertainties.

Multispectral thermal infrared data (TIR: 8-12 micron) are widely used to produce a variety of critical long-term science data records such as Land Surface Temperature and Emissivity (LST&E), and Evapotranspiration (ET). The ECOSTRESS TIR mission launched in mid-2018, and upcoming TIR missions including LSTM, TRISHNA, and the NASA Surface Biology and Geology (SBG) in 2025-2028 will bring a golden age of high spatial resolution (< 100 m), multispectral TIR data, with a potential for a twice-daily global revisit. NASA’s SBG will include both a TIR and VSWIR instrument and is a core component of NASA's new Earth System Observatory (ESO) to improve our understanding of vegetation processes, aquatic ecosystems, urban heat islands and public health, snow/ice, and volcanic activity. In this study we explore the use of ECOSTRESS TIR data in urban heat science and applications. Rapid 21st century urbanization combined with anthropogenic climate warming are significantly increasing heat-related health threats in cities worldwide, and partnerships between city policymakers and scientists are becoming more important as the need to provide data-driven recommendations for sustainability and mitigation efforts becomes critical. We will highlight the use of ECOSTRESS TIR data in monitoring the variability of intra-urban heat islands during extreme heat events over the diurnal cycle, for pinpointing hotspot locations in cities to optimize urban heat mitigation interventions such as cool roofs, cool pavements, cooling centers, and urban greening; and to better understand the thermal properties of urban man-made materials relative to the urban biosphere.

Landsat-9, launched on September 27, 2021, and Landsat-8, launched on February 11, 2013, both carry on-board versions of the Thermal Infrared Sensors (TIRS). The TIRS instruments are very close copies of each other; two spectral bands, pushbroom sensors with three Sensor Chip Assemblies (SCAs) that cover the 15-degree field-of-view. Each spacecraft has a 16-day revisit time, and the two are placed in orbits eight days offset from each other. Modifications were made to Landsat-9 TIRS-2 to upgrade it to a Class-B mission, meaning it has additional redundancies. Also, baffling was added to the Landsat-9 TIRS-2 telescope to mitigate the stray light issue that has plagued Landsat-8 TIRS.

The radiometric performance of the TIRS instruments is monitored using the on-board variable temperature blackbody and views of deep space. Maneuvers to look at and around the moon have provided an assessment of the stray light. The absolute calibration is monitored by vicarious calibration teams at NASA/Jet Propulsion Lab and the Rochester Institute of Technology.

The responsivity of the Landsat-8 TIRS instrument has been degrading over time since November 2020. An apparent contaminant is slowly building up on the focal plane changing the detector’s responsivity nonuniformly across the focal plane. The responsivity has dropped by ~2.5% in Band 10 and ~5% in Band 11, though through updates to the calibration parameters, the image products should remain calibrated to within 0.5%.

Landsat-9 completed a three-month commissioning phase in January 2022. The radiometric performance and initial absolute calibration were assessed. The instrument stability is monitored with the blackbody data over multiple time frames. The noise performance of the instrument is monitored using blackbody data at multiple set-point temperatures.

This paper will cover the recent radiometric performance assessments for both the Landsat-8 TIRS and Landsat-9 TIRS-2 instruments.

Viewing and illumination geometry are known to have significant impact on the remotely sensed retrieval of land surface temperature (LST). Differences appear greatest for areas with mixed components contributing to the pixel-integrated signal, as well as to the amount of shadowing.

Radiative transfer models have been used to assess and, in some cases, adjust for these directional effects on remotely sensed LST, with the aim typically of delivering direction-independent equivalent values. However, use of such models in many cases remains under-evaluated against in-situ data, due in part to the difficulties of retrieving data for the different components in a scene at a variety of different viewing and illumination geometries over a time period where the real surface temperature and sun-sensor geometries are invariant. With LST now classified as an Essential Climate Variable, it is imperative further work is done to ensure these directional effects are well understood and where possible accurately accounted for, particularly when considering any future satellite mission design (e.g. LSTM, SBG, TRISHNA).

To address this issue, a joint ESA-NASA funded airborne campaign (SwathSense) was conducted in summer 2021 – focused on collecting a unique multi-geometry set of airborne and in-situ data over agricultural and urban sites in the UK and Spain.

In the UK, NASA-JPL’s long-wave infrared (LWIR) state-of-the-art Hyperspectral Thermal Emission Spectrometer (HyTES) was flown on a UK research aircraft alongside Specim's FENIX 1K visible and shortwave infrared (VIS-SWIR) hyperspectral imager. In Spain, flights were conducted over agricultural regions with the LWIR hyperspectral TASI imager and VIS-SWIR CASI imager in collaboration with the LIAISE (Land surface Interactions with the Atmosphere over the Iberian Semi-arid Environment) campaign. In-situ field measurements from over fifty sensors – including unmanned aerial vehicles (UAVs) and a multi-angle goniometer equipped with a wide-angle field-of-view thermal camera – were collected to enable assessment of the remote sensing LST retrievals and to understand the extent of any real LST change over the period of the airborne data collections.

We provide an overview of the SwathSense campaign, review the current results from the HyTES, TASI and in-situ sensor analysis, as well as preliminary analysis from the campaign and the implications that these may have for future satellite missions.

Satellite measurements of thermal infrared emission (TIR) provide key diagnostics into the water use of crops and natural ecosystems. Plant growth goes hand-in-hand with the loss of water through the stomata as the leaves take in carbon dioxide. When this water evaporates the required energy is drawn from the leaf surface resulting in a cooling effect and latent heat flux into the surrounding air. Together with evaporation of surface soil water this forms the main terrestrial source of latent heat flux into the atmosphere.

Various algorithms have been developed to exploit the measured thermal impact on land surface temperature and provide diagnostic estimates of the land evaporation (E). Besides TIR they typically require basic visible and near-infrared reflectance to factor in the vegetation coverage and partition E in its source components. At the continental scale these products provide modelers with constraints to the water and energy cycle in an area with high uncertainty in earth system models: at the interface between surface and atmospheric models. Provided the spatial resolution is sufficiently high (100 m or better), the resulting products can provide managers early information on crop condition and overall ecosystem health, or a means to systematically monitor crop water use over large domains. It is especially this last application of TIR observations that is the most demanding in terms of both spatial and temporal resolution.

Fortunately, the availability of high-resolution thermal imagers to capture croplands functional responses is slated to improve dramatically in the coming years thanks to continued Landsat and Sentinel programs, as well as new missions like NASA’s Surface Biology and Geology (SBG) and ESA’s Copernicus Land Surface Temperature Monitoring (LSTM). With this prospect in mind, the science community will benefit from integrating these TIR observations into harmonized land surface temperature products. This requires fundamental understanding of the characteristics of each instrument and on-orbit performance.

This presentation will focus on research to support the adoption of thermal infrared imagery for providing improved monitoring of water use, drought resilience, and its integration within hydrological model frameworks. We will present analysis of the spatial resolving power of current thermal imagers and implications for temperature and evaporation retrieval. Bridges were found to provide a sufficient thermal contrast with the water surface to quantify the line-spread function of thermal imaging systems. The full-width-at-half-max of a gaussian beam model fitted to this transect quantifies the on-orbit spatial resolution of different imagers. This method is used to characterize the spatial resolution of the thermal bands of Landsat (7, 8, and 9) and the ECOsystem Spaceborne Thermal Radiometer Experiment on Space Station (ECOSTRESS). The analysis also investigates spatial sample size growth vesus scan angle. The goal of this research is to facilitate an improved fusion of current and future satellite observations into harmonized products with superior temporal and spatial characteristics.