Contact Us  |  Search  
The National Academies of Sciences, Engineering and Medicine
Partnerships for Enhanced Engagement in Research
Development, Security, and Cooperation
Policy and Global Affairs
Home About Us For Applicants For Grant Recipients Funded Projects Email Updates

Partnerships for enhanced engagement in research (PEER)
Cycle 4 (2015 Deadline)

Satellite-based estimations of river discharge into the Cartagena Bay, Caribbean Colombia: Capacity building to mitigate sources of upstream runoff and associated risks of pollution

PI: Juan D. Restrepo (, Universidad EAFIT
U.S. Partner: Robert Brakenridge, University of Colorado
Project Dates: October 2015 - October 2019

Project Overview

4-70 Q4 2016 RestrepoJuan Restrepo (Project PI, center) during the post
Dr. Restrepo (center) during the poster session at AGU Fall 2016 with U.S. partners Albert Kettner (left), Associate Director of the Darmouth Flood Observatory-UC and Professor James Syvitski, Director of the CSDMS at University of Colorado

The city of Cartagena and its bay, a UNESCO World Heritage Site, have a limited capacity for water resource management, as evident in the degradation of coastal water quality. The greatest source of pollution to the bay and the adjacent Rosario Islands National Marine Park is the Dique Canal, which drains 7% of the country’s largest river, the Magdalena. During the last six decades, coastal fisheries and ecosystems (sea grasses and coral reefs, for example) have almost disappeared, and Cartagena Bay is considered one of the Caribbean Region´s “hot spots” of pollution due to human-induced stressors impacting water quality and ecosystems services. Colombia and its national public environmental institutions have not previously used near real-time satellite observing systems, but this PEER project is aimed at generating near real-time satellite river data for the region by deriving time series of streamflow and flooding extent data. The main goals include (1) estimating river discharge from the Dique Canal-Magdalena River system into Cartagena Bay by applying satellite-derived measurements; (2) relating the constructed river flow series since 1998 to identify climate variability and human impacts for the Magdalena River basin; (3) making satellite-based data available as GIS files so they can be incorporated into end-user decision-support systems; and (4) developing capacity building for implementing satellite-based river freshwater estimations in environmental decision making in Cartagena Bay.

Findings on Cartagena Bay’s pollution and its current inadequacy for recreational use and fish consumption are widely known within the scientific community. In addition, lack of reliable data on freshwater discharge flowing into the bay has limited detailed modeling and applied science on how river fluxes impact physical and chemical characteristics of coastal waters. The ambiguity of the problem means that mitigation strategies are lacking. This PEER project team aims to have a significant impact by generating a clearer and more precise foundation of knowledge on continental runoff fluxes and related marine pollution problems, which could thus inform the development of adaptive strategies at both the community and political level. The planned collaboration between the Dartmouth Flood Observatory at the University of Colorado and EAFIT University-Colombia, the latter with its current project in the region, BASIC-Cartagena (Basin Sea Interaction with Communities), will serve as a joint effort in strengthening the capacity of regional stakeholders to integrate geospatial and satellite products into development and decision making, with the final goal of improving environmental management for Cartagena Bay.

Scientific conclusions at the onset of the fourth year of the project are as follows:

(1) For Andean rivers of Colombia, characterized high rainfall rates, steep relief and limited alluvial plains, microwave satellite radiometry is a robust tool for estimating near-real-time water discharge. In this study, stream flow series during the 1998-2017 were obtained at 14 sites along the Magdalena River and its main tributary, the Cauca.
(2) Observed series of river discharge at the upper basin stations exhibited low seasonality and smaller variations in the satellite signal. In the middle and lower reaches of the Magdalena River, the accuracy of the satellite results is improved at longer time scales. Floods were well recognized by the observed series of river discharge, especially during La Niña event in 2010-2011, the strongest flooding event on record in Colombia. Six out of 14 studied stations witnessed high peak discharges during this period, with stream flow magnitudes being more accurate than the water discharge values gauged at ground-based stations.
(3) The analysis of river-floodplain connectivity, allowed the team to present the following conclusions:
• Once states of flood-connection are identified, turbid waters and disconnection, by analyzing daily MODIS images + water level, we obtained water level thresholds at different connectivity stages.
• After analyzing frequency plots or histograms of water level and DFO water discharge (Q), once the identified thresholds of water level and DFO Q were coupled, we estimated the %s of connectivity time.
• Looking at the statistics, the %s of connectivity time identified by water level and DFO Q are in agreement, in other words, are of the same magnitude! This fact means, that after taking some cautions, DFO can be used to monitor stages of connectivity in different geomorphologic patterns of the Magdalena floodplain, one a braided channel floodplain (Barbacoas lagoon) and the other an anastomosing avulsive floodplain (Momposina depression).

(4) This study is an innovative applied research generating near-real-time satellite river data for the region by deriving time series of stream flow and flooding data back to 1998.

Evidence to Action activities: the project team convened a workshop on disseminating water related information targeted to a select group of policymakers, decision makers, end users, and stakeholders from Colombia and Latin America. This workshop took place in EAFIT University on March 21-22, 2018, as a joint effort of the supported DFO-EAFIT PEER project and the project of the Dartmouth Flood Observatory and the GeoSUR Program, “Mapping of Flood Events and Estimation of River Flow for Latin America and the Caribbean”. The
DFO-GeoSUR project has produced a displaying platform of data for Surface Water Record for Latin America and the Caribbean. By joining both projects efforts, and therefore the different water related data products, results and experiences, the convening workshop discussed with participants how to better understand and use the DFO platform, as well as ingest some of the DFO data directly into their local GIS and decision making analyses. ask feedback on how these data products can be improved and made more accessible to the users. During the meeting, the following water related measurement techniques were discussed in detail: (1) Data availability (data provided by hydrological agencies; satellite based gauging stations DFO; dissemination of water related datasets (DFO & GeoSUR); historical data; (2) 
Use of data (Use cases; the need for data; limitations (temporal / spatial); (3) Information Gaps (potential data poor areas; limitations in sharing data (e.g. Trans Border River systems); desired water related products; study cases in Latin America
All presentation can be downloaded at

Secured additional funding: as a result of conducting this PEER project, Dr. Restrepo obtained a new 3-year grant ($1,250,000.00) from the International Development Research Center IDRC-Canada for the Cartagena Bay (Caribbean Colombia) Project BASIC-Basin Sea Interaction with Communities. The main outcome of this PEER project, which is the development of satellite derived data of water discharge of the Magdalena River in near real time, is the main tool and data platform for the warning system of water pollution that will be developed in the IDRC project in Cartagena Bay.

Potential Development Impacts: In most South America drainage basins, flood events affect many regions and cause huge damage in mainly low-income communities. Environmental decision-making and mitigation strategies for floods frequently lack reliable tools and long enough data records for sustainable drainage basin planning. And flood detection is mostly established on ground-based gauging station records, which could have severe limitations in calibration, accuracy and data availability. This as for example during a flooding event, water can overtop or destroy ground based stations such that the peak of a flood event is not recorded and therefore flood return frequencies are at best guessed. During this EAFIT-DFO-PEER project, the team developed near-real-time satellite-derived estimations of river discharge towards technical capacity building for evaluation of flood magnitudes frequencies and variability along the Magdalena River, Colombia by using the Dartmouth Flood Observatory (DFO) data platform. The data is also available through the GeoSUR project, a collaboratio between DFO and the Latin American Development Bank, CAV. Additional data and maps on flood extent from a local to regional scale for the entire South American rivers were made available through the DFO data portal or the CAV supported geospatial platform GeoSUR. During this project, we have also analyzed the usability and constraints of these water related data products and ask feedback on how these data products can be improved and made more accessible to the users.
In addition, regions in Colombia such as the lower course of the Magdalena have been strongly affected by floods and associated sedimentary fluxes during the last decades. They do not have reliable ground-based data on river discharge and magnitudes of floods. This study is an innovative applied research generating near-real-time satellite river data for the region by deriving time series of stream flow and flooding data back to 1998.
Finally, land-use planning and mitigation strategies in the Magdalena basin require capacity building. Paleo-data on floodplain formation, long-term biophysical data, as well as long-term research to understand the hydrological dynamics, are critical scientific needs towards decision-making. The real limits of the flooding areas and how they are linked to ecosystems services and other relevant processes such as fish migration and recruitment for artisan fisheries are also crucial data. Under the observed increasing trends of land cover change in the Magdalena floodplains, environmental scientists and decision makers face new scientific challenges: (i) obtain the capacity building to monitor and manage floodplain wetlands, (ii) recognize their ecosystem services and value, and (iii) establish the science needs towards its sustainable hydrologic and biologic functioning. Certainly, the DFO data provided by this PEER project in terms of river floodplain connectivity are the main tools to preserve and manage ecosystem services of the Magdalena River and its floodplain lakes.

Stream flow series (water discharge) and frequency flooding maps during the 1998-2018 are now available at 14 sites along the Magdalena River and its main tributary, the Cauca, through the Darmouth Flood Observatory-DFO system at

Environmental and river authorities in Colombia like IDEAM and CORMAGDALENA, and other users such as non-governmental organizations (The Nature Conservancy, Association of the Magdalena River Fishermen) and the scientific community, have access to high quality and near real time water flow data.

Scientific challenges and future plans: the last phase of the project was focused on developing an understanding of floodplain connectivity in the Magdalena River. Since the project team have the whole project objectives already covered and major questions answered, they will focus on writing and submitting the last international peer-review paper titled “Applying satellite data to monitor water discharge and riverfloodplain connectivity in the northern Andes: The Magdalena River, Colombia”. Further challenges include: (1) Major international exposure of the project in some relevant scientific meetings; (2) Strength the network of South American scientists working on river flood detection, which was already initiated during the supplemental activity (Workshop on Floods in South American Rivers, held in EAFIT on March 2018); (3) Carry out final field trips to recognize other major floodplain systems already analyzed with satellite data, but never visited in the field; and (4) During the Evidence-to-Action funded international meeting held in EAFIT, an idea was proposed to develop a course of American Rivers, which will be taught by professors from our current PEER network (USA Universities, Colorado and Montana, EAFIT University in Colombia, and Federal University of the Amazon in Brasil) to students un USA, Colombia and Brazil. Once the course is designed, further funds will be acquired from our universities and other international institutions, but the course philosophy and content will be part of this PEER project results.
4-70 Oct30 Nov3 field trip_Restrepo4-70 Oct30 Nov3 field trip_ _Measuring water discharge
October 30-November 3, 2017 field trip: team members in the boat and in the Barbacoas lagoon.  Measuring water discharge in the Magdalena River [Photos courtesy of Dr. Restrepo]



Back to PEER Cycle 4 Grant Recipients