Remote Sensing in support of Groundwater Studies

Shusen Wang, Richard Fernandes, Rasim Latifovic, Francois Charbonneau, Jianliang Huang, Francis Canisius, Matthew Maloley, Darren Pouliot, Sylvain Leblanc, Jixin Wang, Stéphane Chalifoux

Canada Centre for Remote Sensing, Natural Resources Canada

Groundwater provides drinking water to about one third of all Canadians and up to 80% of the rural population. The protection of clean water supplies has been identified as a national priority. This project addresses the government needs to better understand the dynamics and vulnerability of groundwater resources through developing remote sensing-based modelling tools and data products for key regional aquifers.


The main objective of this project is to support the ESS Groundwater Geoscience Program, which maps groundwater resources, assesses key regional aquifers in Canada and manages and disseminates groundwater information. Specifically, this project focuses on

  1. Aquifer characterization, which includes surface biophysical parameters mapping, soil hydraulic parameter and soil moisture mapping, and aquifer specific yield mapping;
  2. Groundwater dynamics and surface water-groundwater interactions, which includes groundwater recharge mapping and assessment, water budget and seasonal change quantification, and ecosystem and climate change impacts and feedbacks;
  3. Model development and calibration/validation for assessing the aquifer water dynamics using remote sensing data.


Our methods focus on developing remote sensing-based tools for mapping surface hydrological-related parameters of vegetation and soil, as well as models for simulating groundwater-surface interactions under changing environmental conditions. In situ gravity measurements and GRACE satellite data are also used to retrieve total water storage change. The EALCO model developed at Canada Centre for Remote Sensing is used to assimilate the above remote sensing products in assessing aquifer groundwater dynamics.


The project team members and partners have been continuing research and development on methods and prototype products in three areas:

  1. Surface parameters retrieval and validation. A wide range of satellite sensors are used to retrieve hydrological-related parameters of vegetation and soil at different scales, which includes Landsat TM, AVHRR, MODIS, VEGETATION, ENVISAT-ASAR and MERIS, RADARSAT-1 and -2. The retrieved parameters related to vegetation include land cover, land use, and leaf area index (LAI). Figure 1 shows the LAI retrieved from MERIS observations over Southern Ontario (Canisius et al. 2010). The retrieved parameters related to soil include soil texture (, permeability, and soil water content from both optical and Radar satellite sensors. Figure 2 shows an example of soil moisture product estimated using RADARSAT-2’s fully polarimetric data and dual polarization over Chateauguay, QC. Validations and quality control of the products are conducted through field campaign for each aquifer. These parameters contribute to the water modelling and aquifer assessment.
  2. Total water storage change quantification using in situ and satellite gravity measurements. Local scale total water storage change is studied through micro-gravity survey using absolute and relative gravitometers over selected aquifers. National scale total water storage change is studied using the GRACE satellite measurement. Figure 3 shows the total water storage changes of North America over 2007. These studies contribute to the groundwater storage analysis through integrating with soil water studies using both field observation and model simulation in the project.
  3. Model and algorithm development for groundwater assessment through integrating the above remote sensing products and other ancillary data. The EALCO model (Ecological Assimilation of Land and Climate Observations ) developed at CCRS assimilates the EO products mentioned above to simulate the water cycle and assess aquifer recharge (Figure 4). The EALCO model includes dynamic coupling of surface radiation, energy, water, carbon and nitrogen cycles. Its mechanistic representation of the surface physical, physiological, and biogeochemical processes enables users to study the many impacts and feedbacks of climate and ecosystem management on surface water and groundwater interactions. Research is being carried out to enhance the model’s capbility in assimilating new remote sensing products and to calibrate/validate the model using local aquifer parameters (Wang, 2008; 2009; Fernandes et al., 2007), thereby increasing the accuracy of the regional groundwater recharge estimates. EALCO simulates the surface water-groundwater interactions to support assessments of water availability, sustainable yield and vulnerability of regional aquifers under projected climate change and land use scenarios. Current effort of the modelling work is focused on the Waterloo Moraine in southern Ontario.

More details about available data and model products can be obtained by contacting Shusen Wang.

Figure 1: The goal of this figure is to illustrate a Leaf Area Index (LAI) map, differenciating bare ground as low LAI index and ranging to high LAI index for dense canopy forest.

Figure 1: Leaf area index retrieved from MERIS observations over Southern Ontario (03 July 2006).

Larger image

This image shows the estimated soil moisture using Radarsat data. The value varies according to a moisture level of 0.1 to 0.5.

Figure 2: Soil moisture estimated using RADARSAT-2’s fully polarimetric data and dual polarization over Chateauguay, QC. The approach is based on the converging solution of the Integrated Equation Model (IEM) on which the solution to a common roughness is extrapolated. The vegetation effect on soil moisture estimations is obvious due to the multi-scattering effect and attenuation of the scattered signal. With the help of the polarimetric information and signal decomposition technique, it is possible to quantify the backscattered energy generated by the vegetation layer. Reducing the total backscattered intensity by this volumetric contribution results in information that is closely related to the soil surface. From this methodology, it has been possible to generate absolute surface soil moisture maps for every RADARSAT-2 fully polarimetric acquisition

This thematic map shows the standard deviation of changes in water storage. The scale is classified from 0 to 150 mm representing water thickness equivalent.

Figure 3: The Root-Mean-Squares map of the total water storage changes in Water Thickness Equivalent (mm) with respect to the annual mean field derived from GRACE for the period of 12 months in 2007.

This thematic map shows the simulated annual diffuse recharge in the Okanagan Basin. The scale ranges from -500 mm to +400 mm

Figure 4: Annual diffuse recharge simulated over the Okanagan basin using the EALCO model. Note that the effect of groundwater supported evapotranspiration is excluded in this simulation

Larger image



This project involved collaboration with Agriculture and Agri-Food Canada, the Ontario Geological Survey, the Grand River Conservation Authority and the University of Waterloo. The support of the Canadian Space Agency's Government Related Initiatives Program (GRIP) is gratefully acknowledged.


Canisius, F., Fernandes, R. and Chen, J., 2010. Comparison and evaluation of MERIS FR LAI products over mixed land use regions, Remote Sensing of Environment, doi:10.10 16/j.rse.2009.12.010.

Fernandes, R.A., Korolevich, V. and Wang, S. 2007, Trends in land evapotranspiration over Canada for the period 1960-2000 based on in situ climate observations and a land surface model , Journal of Hydrometeorology 8 (5), pp. 1016-1030.

Wang, S., 2008. Simulation of evapotranspiration and its response to plant water and CO2 transfer dynamics. Journal of Hydrometeorology, 9: 426-443, doi: 10.1175/2007JHM918.1.

Wang, S., Yang, Y., Trishchenko, A. P., Barr, A. G., Black, T. A., and McCaughey, H., 2009. Modelling the response of canopy stomatal conductance to humidity. Journal of Hydrometeorology, 10, 521–532, doi: 10.1175/2008JHM1050.1.