Organisation

PhD Defence Mr Juan Francisco Sanchez Moreno

Department of Earth Systems Analysis

Title of Defence

The Power of Rain: Rainfall Variability and Erosion in Cape Verde

Summary

Pressure over land for agriculture and rainfall variations caused by weather and climate variability and by climate change, are increasing soil erosion processes and resulting in desertification.  Tropical semi-arid countries, with occurrence of extreme rainfall events in short time intervals are more susceptible to flash floods and erosion, that are aggravated if the relief effect is added to the equation.  The majority of the semi-arid areas of the world are in developing countries, where pressure over land is even higher and where measures against erosion are still inadequate due to financial limitations.  Data availability poses a challenge for the understanding of rainfall and erosion processes in many semi-arid developing countries.  Cape Verde, off the west coast of Africa, is a semi-arid country conformed by a group of islands located in the tropical zone, subject to a short rainy season with extreme inter annual variation, and where data is scarce.  In Cape Verde, poor agricultural practices along with few arable land and extreme events are intensifying erosion processes leading to desertification.

Parameterization of rainfall in Santiago Island, the largest of Cape Verde, showed that elevation can be used as main predictor for  monthly and seasonal rainfall.   A few days only with extreme high and intense rainfall events can control the rainfall patterns of the month and the season.  On the other hand, extreme weather events act more independently of orography.  For drought studies and agricultural purposes, rainfall variability can be studied in terms of elevation, however erosion models are better studied in terms of single events.  In Cape Verde, rain gauges are not always operative and observations are not all of the same reliability, which makes adequate rainfall parameterization difficult.

As ground rainfall measurements are not always available, estimates from satellite were considered for Cape Verde, particularly from readily available data such as the Tropical Rainfall Meteorological Mission (TRMM) and the Multi Precipitation Estimate from MeteoSat  2nd generation.  Comparison with ground data measurements showed that the products from these sensors underestimate the amount of rainfall.  A newly derived estimate based on cloud top temperature of precipitable clouds correlated better with high temporal resolution ground data measured by an optical disdrometer, and provided a more acceptable fit and rainfall depths within the ranges measured on the ground.  The estimate could be calculated for 15 minutes intervals, which makes it usable as input for erosion studies based on single events.

Rainfall erosivity is an important parameter in erosion studies, as it describes the ability of rainfall to detach soil particles.  An equation to estimate kinetic energy of rainfall was not available for Cape Verde.  Erosivity was expressed in terms of kinetic energy and momentum of rainfall.  The best correlation between rainfall kinetic energy and rainfall intensity was obtained with a power-law equation, and in the form of kinetic energy expenditure in terms of time, in contrast with the usual kinetic energy expressed in terms of rainfall volume.  Momentum was also best correlated with rainfall intensity by a power-law equation, and it was found that momentum and kinetic energy are interchangeable parameters for erosivity estimation.  The new kinetic energy - rainfall intensity equation has been developed and is recommended for erosivity estimation and erosion modeling for Cape Verde, but can be employed in other semi-arid regions with rough topography and at tropical latitudes with similar rainfall characteristics.

The kinetic energy - rainfall intensity relationship was employed to estimate erosivity and to develop relationships correlating it with daily precipitation.  Annual erosivity R-factor was correlated to annual precipitation.  As high temporal resolution rainfall is not available for Cape Verde, an alternative index for erosivity based on monthly precipitation for a long period of time, namely the Modified Fournier Index, was calculated.  The Modified Fournier Index was employed to estimate monthly and long term annual erosivity, and these were mapped and correlated to the R-factor.  Because monthly and seasonal rainfall in Santiago Island are related to elevation, the highest erosivity values occur evidently at higher elevations, coinciding with steep slopes and shallow soils.  Alternative indices for erosivity estimation allow coping with lack of data and permit to take advantage of the available information.

Runoff and erosion was modeled for Ribeira Seca catchment in Santiago Island by openLISEM, a physical event-based hydrological and soil erosion model.  Eight scenarios were created by employing two rainfall events and four types of input maps with different degrees of complexity: low complexity maps consisted in small or large mapping units, while high complexity maps were obtained using geostatistics.  The scenarios provided valid results, however unrealistic calibration parameters were required for the low complexity maps in order to correlate hydrological model output to the measured hydrographs.  Moreover, different complexity representations of input parameters such as soil water storage capacity, result in different runoff and erosion patterns.  Runoff measured at the outlet is not enough as it can be obtained by calibration, whereas looking in the field where the actual runoff and soil losses occur facilitates the selection of the adequate input scenario for the model.

This dissertation as a whole aims to contribute to the knowledge and characterization of rainfall and erosivity for runoff and erosion modeling in a data scarce tropical environment.

Biography

Juan Francisco Sanchez Moreno was born in Bogota, Colombia.  After completing studies in Civil Engineering in the ‘Universidad Nacional de Colombia’, he worked in several private and public organizations, both in Bogota and surrounding areas, and in the Colombian Amazons.  In 2002 he was offered a NFP scholarship to study a Professional Master in Geoinformatics in ITC.  After his return to Colombia he continued working as civil engineer now applying GIS, particularly with the Institute of Urban Development IDU of Bogota.  In 2005 Juan was awarded and AlBan scholarship to study an MSc in Geoinformation Science for Environmental Modeling and Management, offered by four institutions: U. of Southampton, Lund University, University of Warsow, and ITC.  His MSc thesis, ‘Applicability of Knowledge-Based and Fuzzy Theory-Oriented Approaches to Land Suitability for Upland Rice and Rubber, as compared to the Farmers' Perception’ received the 2007 ‘Hissink’ Award from the Dutch Soils Society for best thesis in a soil related topic.  In 2007 he graduated with honors from the MSc program and was offered a PhD position under the DESIRE program against land desertification.  During his studies Juan was member of the ITC PhD committee and the ITC volleyball team.

Currently Juan Francisco lives in Eindhoven, The Netherlands, with his wife Sally and their daughter Sara.

Sanchez Moreno, J.F., Jetten, V.G. (Promotor) and Mannaerts, C.M. (assistant-promotor)  (2012) The power of rain : rainfall variability and erosion in Cape Verde. Enschede, University of Twente Faculty of Geo-Information and Earth Observation (ITC), 2012. ITC Dissertation 217, ISBN: 978-90-6164-343-2.

   Full text

previous page
more events