Congo Wetlands Hydrology Using SAR, InSAR, and Radar Altimetry

Date
2017-05
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract

Extensive wetlands expand along the banks of the Congo river and its tributaries. Estimating wetlands water storages as well as its sources and sinks is of great importance to understand water balances in the fluvial systems and the role of wetlands in nutrient and sediment transport. In this dissertation, Synthetic Aperture Radar (SAR), Interferometry SAR (InSAR), and radar altimetry are deployed to study the spatio-temporal variations of water level changes (∂h⁄∂t) and water storages, and the influxes and effluxes through the wetlands. First, a regression model integrating PALSAR backscattering coefficient ("σ" _0), Envisat altimetry, and MODIS VCF is proposed to estimate ∂h⁄∂t in the Malebo Pool where non-forested and forested vegetations exist. The accuracy of "σ" _0 method is adequate in non-forested regions where InSAR could lose coherence during high water level, whereas the accuracy degrades in dense forested wetlands where InSAR could maintain coherence to measure relative ∂h⁄∂t. Therefore, a stack of PALSAR interferograms are generated to map multi-temporal high spatial resolution ∂h⁄∂t over the forested wetlands near city Lisala. The InSAR measurements show that ∂h⁄∂t is subtle and the water flow is not well confined during low water seasons while ∂h⁄∂t shows greatest gradient perpendicular to river flow direction in high water seasons. Next, a new method is developed to reveal the inter-annual absolute water storages by integrating spatially varying ∂h⁄∂t and time series of water depth from Envisat altimetry. The mean annual amplitude over the studied 7,777 km2 wetlands from 2002 to 2011 is 3.98±0.59 km3, with maximum water volume to be 6.3±0.68 km3 in the wet year of 2002 and minimum volume to be 2.2±0.61 km3 in the dry year of 2005. Finally, mass balance analysis results suggest that local water (i.e. upland runoff and rainfall) contributes more than 85% of total inflow while regional water (i.e. wetland river exchanges) supplies less than 15% of total inflow. Wetland river exchanges contribute 59% to 80% to the total outflow while evapotranspiration contributes 20% to 41% of the total outflow. A wetland model was applied to derive two site specific hydrological parameters, flow conductance (K_f) and depth exponent (β).

Description
Keywords
Congo wetlands, Synthetic aperture radar (SAR), Radar altimetry
Citation
Portions of this document appear in: Yuan, Ting, Hyongki Lee, and Hahn Jung. "Toward estimating wetland water level changes based on hydrological sensitivity analysis of PALSAR backscattering coefficients over different vegetation fields." Remote Sensing 7, no. 3 (2015): 3153-3183. And in: Yuan, Ting, Hyongki Lee, and Hahn Chul Jung. "Congo floodplain hydraulics using PALSAR InSAR and Envisat altimetry data." In Remote Sensing of Hydrological Extremes, pp. 65-81. Springer, Cham, 2017.