Improving distributed hydrologic models using multiscale thermal infrared, near infrared, and visible imagery from sUAS and satellite-based sensors
Can the use of sUAS combined with satellite data effectively improve calibration of distributed hydrologic models? Previous work has focused on developing techniques to control variables such as evapotranspiration using satellite remote sensing. The critical component is the equilibrium surface temperature which is normally calibrated using thermal remote sensing data from MODIS (MODerate resolution Imaging Spectroradiometer) or AATSR (Advanced Along-Track Scanning Radiometer) which have a 1 km ground sample distance. This study seeks to improve thermal calibration techniques using much finer-resolution thermal measurements derived from a sUAS. The use of a sUAS will allow for the collection of high resolution imagery (< 1 cm – 1 m or more depending upon research goals) of landscape components, repeatedly and during desired time frames. Thermal data from a sUAS will provide much greater spatial and temporal resolution than satellite-based measurements, and will be obtained at a relatively low cost. This study seeks to determine the added benefit of improved spatial and temporal resolution of observations, evaluated through the existing model calibration framework of a sub-daily, sub-kilometer hydrologic model with complex terrain and vegetation, typical of many mountain headwaters systems experiencing change in the West. If successful, the techniques developed here can be used to apply an appropriate and cost-effective combination of satellite and/or sUAS observations to improve other hydrologic models of interest to Reclamation.
Need and Benefit
The proposed project directly addresses needs identified by Reclamation's Water Operations and Planning (WP)
Research Program, including to:
•"Explore possible hydrologic model performance enhancements by leveraging new or additional data for model
calibration and forcings" (Water Supply #4): The ability to calibrate existing models is crucial to accurate forecasting.
Current methods employ streamflow and satellite measurements, which do not provide an assessment of the
predictive capabilities for other hydrological states and fluxes. Integrated observations also do not provide insight into
a model's ability to capture spatial patterns within a catchment (Stisen et.al. in press). This simplification can result in
calibration of the wrong parameters and resulting errors in streamflow predictions. High resolution data are particularly
important given the differing energy balance fluxes from bare patches and different plant species (Detto et al. 2006,
Vivoni et al. 2010). sUAS sensors provide the opportunity for calibration datasets with precision not available from
coarser resolution sensors.
•"Develop approaches and tools to support and improve long term planning for changing or uncertain future hydrologic
conditions" (Water Operations Models #1): The changing hydrologic conditions of many watersheds important to
Reclamation, including the unprecedented insect-induced tree die off in the proposed study area in Grand Lake (e.g.
Bearup et al. 2014), necessitate the need for more complex and physically meaningful tools that push beyond
historical relationships. The development of these tools, in turn, requires robust calibration datasets that consider the
full hydrologic cycle, such as the land surface temperature dataset proposed here.
•"Demonstrate the application of new or improved data with existing models toward enhanced operational outcomes"
(Water Operations Models #3): This research will provide important first steps into the combining of new technology
into the broader science of hydrologic modeling. We anticipate continuing the use of sUAS's into a number of
hydrology related topics salient to Reclamation's mission.
•"Investigate new technologies and methods to enhance basin hydrologic condition monitoring data" (Water Supply
#1): The use of sUAS technology is a new and burgeoning technology in many fields. Presently, there are few
examples of the coupling of sUAS and micro sensors for improving hydrologic models. This S&T proposal is the first to
address this specific application of sUAS technology with hydrologic modeling within Reclamation. There are no
existing capabilities within Reclamation and we seek to develop this.
The anticipated benefit of high precision and multi-temporal data provided by sUAS data is the ability to parameterize
and calibrate relatively highly-resolved hydrologic models to improve streamflow estimates and ultimately forecasting
ability. Providing accurate model estimates is crucial to Reclamation's long-term climate predictions and the effect on
the nation's water supply and disbursement. The continued use of sUAS's to provide quick and cheaper surveys over
a broader range of applications, such as water infrastructure monitoring, environmental issues such as water delivery
reliability, water quality, river habitat restoration, water operations such as reservoir capacity studies are all proven
uses. The sUAS experience gained will easily leverage into all of these research and applications and fit in well with
both Reclamation's overall mission and the mission and vision of the S&T Program.
If not funded or acted on now, Reclamation will miss an opportunity to push our modeling efforts to the forefront of the
science. The incorporation of sUAS technology into hydrologic modeling is starting to happen, with evidence of the
increasing use appearing in the peer-reviewed literature; however our proposed application for use in water supply
planning and operatio
Contact the Principal Investigator for information about partners.
Contact the Principal Investigator for information about these documents.