In This Space
Image credit: Servir (USAID/NASA)
Water Quantity & Quality
Water body temperature derived from satellite measurements provide information on the temperature of the water close to the surface, or the water “skin”, but no information on the vertical temperature profile of the water body. To derive water surface temperature, a satellite sensor has to measure in the thermal infrared (TIR - emissive) part of the electromagnetic spectrum (between 8 to 14 µm). EO systems with thermal optical sensors are available, such as Landsat and ASTER. Emissive thermal infra-red images typically have lower spatial resolution than corresponding reflective optical images from the same.
Quantitative estimation of temperatures from TIR images requires compensation for absorption by atmospheric gases (water vapour, ozone and carbon dioxide) and upwelling atmospheric emitted radiance. The most common method to compensate for atmospheric effects is usage of radiative transfer models such as MODTRAN. Thermal infrared temperatures may also be affected by surface effects (such as multiple reflection, and sub‑pixel mixing).
Temperature data obtained from satellite sensors and in situ measurements are desirable to understand the thermal processes taking place in water bodies. Satellite-based monitoring may extend and supplement data that can be obtained from in-situ devices, including automated sensors that may provide data via telemetry. In regions with vast numbers of lakes (e.g., Canada, Alaska) very little prior information may be available in exploration and development areas.
The Sentinel-3 satellite will include the Sea and Land Surface Temperature Radiometer (SLSTR) which is based on Envisat’s (Advanced) Along Track Scanning Radiometers AATSR.
The water body temperature product can deliver temperature calibrated images over lake areas (raster), along with horizontal profile plots and summary reports. Multi-temporal and multi-season characteristics can be evaluated.
Known restrictions / limitations
Temperature calculation may not be possible for small water bodies due to the low spatial resolution of thermal sensors.
Lifecycle stage and demand
All lifecycle stages:
Monitoring of water bodies over time to capture seasonal variation in properties.
Geographic coverage and demand
Coverage and demand is global, but focused on locations where access is challenging, or the sheer number of lakes makes baseline data gathering very challenging.
HC:2102 Understanding hydrogeology
Input data sources
Optical: HR2, MR1 with thermal infrared bands
Water body temperature estimation requires usage of sensors capable of acquiring data in the thermal part of the spectrum, e.g., Landsat-7, Landsat- 8, Landsat- 5 (archive only). ASTER sensors (since 1999) provide 5 TIR bands with 90 m resolution.
Supporting data: In situ measurements
Spatial resolution and coverage
Spatial resolution: 30–90 m.
The best available spatial resolutions are: 60 m from Landsat-7; 100 m (resampled to 30 m) from Landsat-8; and 90 m for ASTER.
Minimum Mapping Unit (MMU)
The minimum mapping unit is constrained by the minimum size of lakes that can be imaged. For example, from Landsat-7 (60 m), lakes can be monitored for average temperature that are greater than approximately 1.4 ha; however, lake position and extent would need to be delineated from higher resolution imagery, to constrain thermal returns to “on-target” areas.
Accuracy / constraints
Landsat-8: 41 meters circular error, 90-percent confidence.
AVHRR: The validation against in situ observations exhibits biases in the range of −0.4–0.6 K.
Sentinel 3 SLSTR: will allow determining global sea-surface temperature with an accuracy of 0.3K.
Accurate discrimination of land from water pixels and mixed pixels is important. Precise geocoding and ortho-rectification is important especially for the small lakes.
Accuracy assessment approach & quality control measures
Comparison with in-situ measurements.
Frequency / timeliness
Observation frequency: Depending on satellite revisit time (e.g., Landsat8, 16 days). Frequency of historical maps is highly variable depending on the archive.
Timeliness of delivery: Processing can be completed in near real time (< 24 hours) depending on availability of base images.
Archive and ESA (Sentinel) data available at no charge. Systematic acquisition of Landsat-8 at no charge.
Archived products, including basic archived ASTER data products are available to all users at no cost. Availability may be limited for specific dates.
Delivery / output format
# of Pages:
Internal – Project consortium and science partners
External – ESA
This page has no comments.