Satellite-based estimates of surface water dynamics in the Congo River Basin

https://doi.org/10.1016/j.jag.2017.11.015Get rights and content

Highlights

  • Satellite-derived Surface Water (SW) dynamics is quantified in Congo River Basin.

  • SW extent shows high interannual variability modulated by IOD/ENSO events.

  • SW storage contributes to 19 ± 5% of the annual terrestrial water storage (TWS).

  • SW storage represents from 33 ± 7% of TWS change in the Middle Congo River Basin.

  • SW storage accounts for 6 ± 2% of the annual water volume that flows to the Atlantic.

Abstract

In the Congo River Basin (CRB), due to the lack of contemporary in situ observations, there is a limited understanding of the large-scale variability of its present-day hydrologic components and their link with climate. In this context, remote sensing observations provide a unique opportunity to better characterize those dynamics. Analyzing the Global Inundation Extent Multi-Satellite (GIEMS) time series, we first show that surface water extent (SWE) exhibits marked seasonal patterns, well distributed along the major rivers and their tributaries, and with two annual maxima located: i) in the lakes region of the Lwalaba sub-basin and ii) in the “Cuvette Centrale”, including Tumba and Mai-Ndombe Lakes. At an interannual time scale, we show that SWE variability is influenced by ENSO and the Indian Ocean dipole events. We then estimate water level maps and surface water storage (SWS) in floodplains, lakes, rivers and wetlands of the CRB, over the period 2003–2007, using a multi-satellite approach, which combines the GIEMS dataset with the water level measurements derived from the ENVISAT altimeter heights. The mean annual variation in SWS in the CRB is 81 ± 24 km3 and contributes to 19 ± 5% of the annual variations of GRACE-derived terrestrial water storage (33 ± 7% in the Middle Congo). It represents also ∼6 ± 2% of the annual water volume that flows from the Congo River into the Atlantic Ocean.

Introduction

Despite its importance, the Congo River Basin (CRB, acronym glossary is given in Appendix A1), located in the central region of Africa, has not attracted as much attention among the climate and hydrology communities as has the Amazon Basin or other large rivers in the world (Alsdorf et al., 2016). Up to now, there is still an insufficient knowledge of the regional hydro-climatic characteristics and changes in this region, even though the CRB plays a crucial role at global and regional scales. Firstly, the CRB is remarkable as the second largest river system of the world in terms of both water discharge, with a mean annual flow of ∼40,600 m3/s, and drainage basin size (∼3.7 × 106 km2) (Laraque et al., 2001, Laraque et al., 2009). It also plays a key role in the Earth system as one of the three main convective centers in the Tropics, with the Amazon River basin and the ‘maritime continent’ of Eastern Indian and western tropical Pacific Oceans (Hastenrath, 1985). Secondly, more than 80% of people in the CRB live exclusively on activities that are highly dependent on climate and water resource availability: fisheries, agriculture and livestock (Bele et al., 2010). In this region, the food production depends heavily on rain-fed agriculture, leading the population particularly vulnerable to food insecurity (Brown et al., 2014). Moreover, a couple of studies have shown that the CRB has already experienced changes in climate variability and in the hydrological system (Mahé and Olivry, 1999; Camberlin et al., 2001; Laraque et al., 2001; Samba et al., 2008; Samba and Nganga, 2012). Thirdly, about 50% of the CRB land area is covered by tropical forest (∼190 106 ha, Verhegghen et al., 2012), representing about 18% of the world’s tropical forests (∼1100 106 ha, Achard et al., 2002), and playing a crucial role as a sink of CO2, storing about 50 billion tons of carbon (Verhegghen et al., 2012). In a recent study, Dargie et al., (2017) highlighted that the “Cuvette Centrale” peatland (Fig. 1) stores about 30 billion tons of carbon. This total amount of carbon is equivalent to ∼80 billion tons of CO2 or about 2.3 years of current global anthropogenic emissions (∼35 billion tons in 2015, Olivier et al., 2016). This stock is particularly vulnerable to land-use change and any future change in the water cycle. For all these reasons, there is an obvious need to better understand the CRB dynamic and to characterize its vulnerability to climate change and other crucial challenges. In particular, it is necessary to gain solid knowledge about the past and current hydro-climate processes of the CRB, in order to significantly reduce the uncertainties associated with future climate response under global warming. The limited understanding of the CRB's hydro-climate processes results mainly from the lack of in situ data availability: the network of stations, which data are publicly released, is sparse and poorly maintained, and it is substantially difficult of perform fieldwork, notably in the swamps. However, recent developments and improvements in remote sensing technology provide more observations than ever before (Alsdorf et al., 2007; Prigent et al., 2016) and allow us the unique opportunity to better understand the spatial and temporal variability of the CRB’s hydro-climatic patterns.

In this study, our primary focus is on the CRB surface water (SW) dynamics, a key component of the land water budget equation. The SW, corresponding to water stored in rivers, lakes, wetlands, floodplains and in man-made reservoirs, is crucial to the survival of all living organisms, including humans and is a precious resource in term of biodiversity, ecology, water management and economy. Moreover, SW storage (SWS) plays a major role at all scales in the terrestrial water balance and in the Earth's climate system variability, through its interactions with the atmosphere and ocean. Up until now, the spatial and temporal dynamics of SW stored on the Earth's surface remains still largely unknown (Alsdorf et al., 2007). Since the last decades, progresses in satellite remote sensing are improving substantially our understanding of SW dynamics in the major river basins of the world. Among these derived-products, radar altimetry is providing since the early 1990s a monitoring of water levels variations of lakes, rivers, floodplains and reservoirs (Birkett, 1995; Crétaux and Birkett, 2006; Calmant et al., 2008). Additionally, it is possible to extract locally the extent of water bodies using satellite imagery, which, combined with altimetry data, enable the SWS estimation of lakes and reservoirs (Baup et al., 2014; Crétaux et al., 2016) and of floodplains (Frappart et al., 2005). More recently, merging information derived from active and passive microwave sensors and from optical data, the Global Inundation Extent from Multi-Satellite (GIEMS) dataset (Prigent et al., 2007; Papa et al., 2010; Prigent et al., 2016) offers unprecedented information on the variations of SW extent (SWE) at the global scale. The combination of GIEMS estimates with radar altimetry observations has further allowed the provision of spatio-temporal variations of SWS in large tropical river basins, such as the Amazon, Ganges–Brahmaputra and Orinoco basins (Frappart et al., 2008, Frappart et al., 2010, Frappart et al., 2012, Frappart et al., 2015b; Papa et al., 2015). Recently, a few studies tried to understand the SW dynamics in the CRB using remote sensing and/or modeling (Rosenqvist and Birkett, 2002; Bwangoy et al., 2010; Jung et al., 2010; Beighley et al., 2011; Lee et al., 2011; Tshimanga et al., 2011; Tshimanga and Hughes, 2012; O’Loughlin et al., 2013; Becker et al., 2014; Betbeder et al., 2014; Lee et al., 2014, Lee et al., 2015). For instance, Rosenqvist and Birkett (2002) demonstrated that Synthetic Aperture Radar (SAR) image mosaics can be used to appraise the maximum extents of flooding in the CRB, but were not relevant to assess the SW dynamics and ranges of the variations. Bwangoy et al., (2010) demonstrated the utility of optical and radar remotely sensed data in characterizing the wetlands of the “Cuvette Centrale”. They estimated that the wetlands cover an area of 32% in the “Cuvette Centrale”, equivalent to 360,000 km2. Crowley et al., (2006), using Gravity Recovery and Climate Experiment (GRACE) data, estimated the terrestrial water storage (surface water storage plus groundwater storage and soil moisture) within the CRB. Over 4 years (2002–2006), the estimate exhibited significant seasonal variations (30 ± 6 mm of equivalent water thickness) and long-term negative trend (∼−70 km3/year). Lee et al., (2011), using GRACE data and other satellite measurements, estimated that the amount of water annually filling and draining the Congo wetlands is about 111 km3, i.e. one-third the magnitude of the water volumes found on the mainstream Amazon floodplain. Lee et al., (2014), integrating terrestrial water storage (TWS) changes from GRACE, water level changes from radar altimetry, and inundation extents from SAR imagery, quantified TWS change and its surface and subsurface components over the central CRB. They showed that annual variations of the TWS changes during the period of 2007–2010 from 21 to 31 km3 and are mostly controlled by surface storage changes. Lee et al., (2015) developed water depth maps over the “Cuvette Centrale” based on a linear regression model from altimetry and imagery data. They reported in their study area water storage volumes of about 11 km3 (Dec-2006), 10 km3 (Dec-2007), and 9 km3 (Dec-2008). Finally, Becker et al., (2014) released an unprecedented dataset of water level time series over the entire CRB for the period 2003–2009, obtained from the ENVISAT radar altimetry mission. From this unique data set, they proposed an altimeter-based river level height regionalization scheme and thus identified nine distinct hydrological regions in the CRB.

Hence, to supplement this work, we analyze the spatio-temporal variability of SW extent and storage in the CRB, at seasonal and interannual time-scales. For this, along with GIEMS data covering the period 1993–2007, we further develop the observation-based technique combining SWE and radar altimetry measurements (Frappart et al., 2008) to estimate the CRB’s SWS variations over the period 2003–2007. The results are evaluated and analyzed along with other in situ and remote sensing measurements of three hydrological parameters (discharge, rainfall and terrestrial water storage). The comparisons with the latter will provide, for the first time, the time series of both SW and sub-surface water (SSW) variations distributed throughout the CRB. The paper is structured as follows. In Section 2 we briefly describe the CRB. Section 3 presents the datasets used in this study. In Section 4, we analyze the SWE dynamics from the GIEMS dataset at both seasonal and interannual time-scales for the period 1993–2007. Section 5 describes the methodology for deriving SWS from a combination of multi-satellite observations and the results are presented and discussed over the 2003–2007 period. Finally, conclusions and perspectives are presented in Section 6.

Section snippets

The study region: The Congo River Basin (CRB)

The CRB is a transboundary basin located in equatorial Africa (Fig. 1). In the heart of the CRB, the shallow depression along the equator is named the “Cuvette Centrale” (Bernard, 1945) (Fig. 1). The Congo River begins its course at the Chambeshi River (Fig. 2), rising south of the Lake Tanganyika and transferred by the Bangweulu Swamps. After flowing through Lake Mweru, it joins the Lwalaba River (Balek, 1977). The permanent surface area of Lake Bangweulu is about 3000 km2 and can expand to

ENVISAT radar altimeter observations

The ENVIronmental SATellite (ENVISAT) was launched in March 2002 by the European Space Agency and its mission was ended in April 2012. The ENVISAT mission carried, among others instruments, a nadir radar altimeter (Wehr and Attema, 2001). Along its ground tracks, repeated every 35 days, we can extract a water level time series, or “virtual stations” (VS), at each intersection with wetlands, large rivers and smaller tributaries of the CRB. The raw ENVISAT data are freely distributed by the

Spatio-temporal variations of surface water extent (SWE)

The mean and maximum of the SWE per grid cell over 1993–2007 are displayed in Fig. 3a and b. A very realistic spatial distribution of the major rivers (Congo, Kasai, Ubangi and Lwalaba) and some tributaries is shown by these maps. Associated inundated areas, wetlands and the region of the “Cuvette Centrale” are also well delineated. Maxima are located in two regions: (i) in the Lwalaba basin, mainly along the major lakes; and (ii) in the “Cuvette Centrale” along the Congo main stream, between

Spatio-temporal variations of surface water storage (SWS)

Here we present the spatio-temporal variability of SWS estimated by combining the SWE from GIEMS with the 350 altimeter-derived water level heights. The two-step methodology is described briefly in the following sections and we refer to Frappart et al., 2008, Frappart et al., 2011 for more details. The results are analyzed for 2003–2007, the common period of availability for both datasets.

Conclusions and perspectives

This work presents an unparalleled analysis of the dynamics of surface water extent (1993–2007) and storage (2003–2007) in the CRB. First, we show that the SWE seasonal patterns from GIEMS dataset exhibit very realistic distributions of major rivers (Congo, Kasai, Ubangi and Lwalaba) and their tributaries, with two maxima located: i) in the lake region of the Lwalaba sub-basin and ii) in the “Cuvette Centrale”, including Tumba and Mai-Ndombe Lakes. For the period 1993–2007, we found a ENSO/IOD

Acknowledgments

This study was funded by CNES TOSCA project “Monitoring of terrestrial water storage variability in the Tropics. An integrated approach of multi-satellite observations, in situ measurements and modeling (HyVarMultiObs)” (2012–2015).

References (97)

  • H. Lee et al.

    Mapping wetland water depths over the central Congo Basin using PALSAR ScanSAR, Envisat altimetry, and MODIS VCF data

    Remote Sens. Environ.

    (2015)
  • G. Mahé et al.

    Assessment of freshwater yields to the ocean along the intertropical Atlantic coast of Africa (1951–1989)

    Comptes Rendus de léAcad’mie des Sciences-Series IIA-Earth and Planetary Science

    (1999)
  • G. Ramillien et al.

    Time variations of land water storage from an inversion of 2 years of GRACE geoids

    Earth Planet. Sci. Lett.

    (2005)
  • J.S. Santos Da Silva et al.

    Water levels in the Amazon basin derived from the ERS 2 and ENVISAT radar altimetry missions

    Remote Sens. Environ.

    (2010)
  • R.M. Tshimanga et al.

    Climate change and impacts on the hydrology of the Congo Basin: the case of the northern sub-basins of the Oubangui and Sangha Rivers

    Phys. Chem. Earth Parts A/B/C

    (2012)
  • R.M. Tshimanga et al.

    Initial calibration of a semi-distributed rainfall runoff model for the Congo River basin

    Phys. Chem. Earth Parts A/B/C

    (2011)
  • T. Wehr et al.

    Geophysical validation of ENVISAT data products

    Adv. Space Res.

    (2001)
  • F. Achard et al.

    Determination of deforestation rates of the world’s humid tropical forests

    Science

    (2002)
  • R.F. Adler

    The version-2 global precipitation climatology project (GPCP) monthly precipitation analysis (1979–present)

    J. Hydrometeorol.

    (2003)
  • D.E. Alsdorf et al.

    Measuring surface water from space

    Rev. Geophys.

    (2007)
  • D. Alsdorf et al.

    Opportunities for hydrologic research in the Congo Basin

    Rev. Geophys.

    (2016)
  • N. Balas et al.

    The relationship of rainfall variability in West Central Africa to sea-surface temperature fluctuations

    Int. J. Climatol.

    (2007)
  • J. Balek

    Hydrology and Water Resources in Tropical Africa

    (1977)
  • J.L. Bamber

    Ice sheet altimeter processing scheme

    Int. J. Remote Sens.

    (1994)
  • F. Baup et al.

    Combining high-resolution satellite images and altimetry to estimate the volume of small lakes

    Hydrol. Earth Syst. Sci.

    (2014)
  • L.C. Beadle

    The Inland Waters of Africa. An Introduction to Tropical Limnology

    (1981)
  • M. Becker et al.

    Water level fluctuations in the Congo Basin derived from ENVISAT satellite altimetry

    Remote Sens.

    (2014)
  • S.K. Behera et al.

    Subtropical SST dipole events in the southern Indian Ocean

    Geophys. Res. Lett.

    (2001)
  • R.E. Beighley et al.

    Comparing satellite derived precipitation datasets using the Hillslope River Routing (HRR) model in the Congo River Basin

    Hydrol. Processes

    (2011)
  • Y. Bele et al.

    Central Africa: the Effects of Climate Change in the Congo Basin; the Need to Support Local Adaptive Capacity

    (2010)
  • É. Bernard

    Le Climat écologique: De La Cuvette Centrale Congolaise

    (1945)
  • J. Betbeder et al.

    Mapping of central africa forested wetlands using remote sensing

    IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens.

    (2014)
  • C. Birkett et al.

    Indian Ocean climate event brings floods to East Africa’s lakes and the Sudd Marsh

    Geophys. Res. Lett.

    (1999)
  • C.M. Birkett

    The contribution of TOPEX/POSEIDON to the global monitoring of climatically sensitive lakes

    J. Geophys. Res.: Oceans

    (1995)
  • P. Bousquet

    Contribution of anthropogenic and natural sources to atmospheric methane variability

    Nature

    (2006)
  • J.-P. Bricquet

    Les écoulements du Congo à Brazzaville et la spatialisation des apports

  • H.C.P. Brown et al.

    Climate change and forest communities: prospects for building institutional adaptive capacity in the Congo Basin forests

    Ambio

    (2014)
  • F. Bultot

    Atlas climatique du bassin congolais

    (1971)
  • W. Cai et al.

    Increased frequency of extreme Indian Ocean Dipole events due to greenhouse warming

    Nature

    (2014)
  • S. Calmant et al.

    Monitoring continental surface waters by satellite altimetry

    Surv. Geophys.

    (2008)
  • P. Camberlin et al.

    Seasonality and atmospheric dynamics of the teleconnection between African rainfall and tropical sea-surface temperature: Atlantic vs. ENSO

    Int. J. Climatol.

    (2001)
  • D. Conway et al.

    Rainfall variability in East Africa: implications for natural resources management and livelihoods

    Philos. Trans. R. Soc. Lond. A: Math. Phys. Eng. Sci.

    (2005)
  • J.-F. Crétaux et al.

    Lakes studies from satellite altimetry

    Coastal Altimetry

    (2011)
  • J.-F. Crétaux et al.

    Lake volume monitoring from space

    Surv. Geophys.

    (2016)
  • J.W. Crowley et al.

    Land water storage within the Congo Basin inferred from GRACE satellite gravity data

    Geophys. Res. Lett.

    (2006)
  • G.C. Dargie et al.

    Age, extent and carbon storage of the central Congo Basin peatland complex

    Nature

    (2017)
  • B. Decharme et al.

    A new river flooding scheme for global climate applications: off-line evaluation over South America

    J. Geophys. Res.

    (2008)
  • B. Decharme et al.

    Global off-line evaluation of the ISBA-TRIP flood model

    Clim. Dyn.

    (2011)
  • Cited by (52)

    • Assessing the performance of the Tiangong-2 wide-swath imaging altimeter observations for water level monitoring over complex and shallow lakes

      2022, Journal of Hydrology
      Citation Excerpt :

      Surface water refers to water retained in rivers, streams, lakes, reservoirs, and wetlands, and accounts for less than 1% of the total amount of water on Earth (Shiklomanov, 1993; Lehner and Doll, 2004). However, it has a large impact on the hydrological and biogeochemical cycles (Strayer and Dudgeon, 2010; Becker et al., 2018; Melo and Getirana, 2019). As a significant hydrological parameter, the water level influences the transfer of water and energy between atmosphere and the land/water surface (Vassiljev et al., 1995; Vincent, 2009).

    View all citing articles on Scopus
    View full text