Transcript
Journal of Environmental Management 88 (2008) 1131–1140 Developing a GIS based integrated approach to flood managementin Trinidad, West Indies Bheshem Ramlal a , Serwan M.J. Baban b, a The Centre for Caribbean Land and Environmental Appraisal Research (CLEAR), The Office of Research, The University of the West Indies,St. Augustine, Trinidad and Tobago b The Centre for Geoinformatics Research and Environmental Assessment Technology (GREAT), School of Environmental Science and Management,Southern Cross University, P.O. Box 157, Lismore, NSW 2480, Australia Received 30 April 2005; received in revised form 29 January 2007; accepted 6 June 2007Available online 21 August 2007 Abstract Trinidad and Tobago is plagued with a perennial flooding problem. The higher levels of rainfall in the wet season often lead toextensive flooding in the low-lying areas of the country. This has lead to significant damage to livestock, agricultural produce, homes andbusinesses particularly in the Caparo River Basin. Clearly, there is a need for developing flood mitigation and management strategies tomanage flooding in the areas most affected. This paper utilizes geographic information systems to map the extent of the flooding,estimate soil loss due to erosion and estimate sediment loading in the rivers in the Caparo River Basin. In addition, the project requiredthe development of a watershed management plan and a flood control plan. The results indicate that flooding was caused by severalfactors including clear cutting of vegetative cover, especially in areas of steep slopes that lead to sediment filled rivers and narrowwaterways. Other factors include poor agricultural practices, and uncontrolled development in floodplains. Recommendations tomanage floods in the Caparo River Basin have been provided. r 2007 Published by Elsevier Ltd. Keywords: Flood management; GIS; Trinidad 1. Introduction Flooding is the most common of disastrous acts of nature among all catastrophes leading to economic lossesand death (Sharma and Priya, 2001). A flood can bedefined as a mass of water, which produces runoff on landthat is not normally covered by water or alternatively aflood is a fairly high flow, which overburdens the naturalchannel, provided for the runoff (Ward, 1978).Flooding is a major problem in Trinidad and Tobago. Inrecent years especially, there have been many severeflooding events which have led to significant damage tolivestock, agricultural produce, homes and businesses(Bryce, 1999; Baban and Kantarsingh, 2005). The problem stems from the inappropriate use of landsthat are vulnerable to erosion, quick water runoff and slopefailure. Poor land use practices include slash and burnagriculture, quarrying, illegal logging, forest fires, andillegal settlements. These have lead to the soil becomingmore exposed and therefore more susceptible to beingwashed away during periods of heavy rainfall andsubsequent runoff (Few et al., 2004). Consequently, heavysedimentation occurs in the river channels causing thesechannels to be reduced in size. This ultimately leads toflooding. In addition, the lack of vegetative cover leads tomuch shorter lag times between rainfall and the waterreaching the waterways causing the already reducedchannels to overflow leading to massive floods (Dion,2002). Inappropriate land use is further aggravated byseveral problems including the absence of relevant infor-mation on the level of vulnerability of different areas todamage due to these usages, the lack of political will toaddress the problems and the lack of adequate resources ARTICLE IN PRESS www.elsevier.com/locate/jenvman0301-4797/$-see front matter r 2007 Published by Elsevier Ltd.doi:10.1016/j.jenvman.2007.06.010 Corresponding author. E-mail addresses:
[email protected] (B. Ramlal),
[email protected] (S.M.J. Baban). for the relevant authorities to enforce the laws of thecountry (Smart, 2004).With global climate change and the expected increase inextreme climatic events, the possibility of severe dry spellsand heavy rainstorms is also likely to further exacerbate thesituation, especially in the most vulnerable areas. There istherefore an urgent need to introduce mitigation measuresto ensure that these areas are protected so that erosion andflooding is minimized. The first step in achieving this is toidentify the nature and extent of vulnerability of the areasunder consideration. Next, determine the most appropriatemitigation measures that should be used to address theproblem. Finally, these measures must be implemented andmaintained.One of the best possible approaches for identifyingvulnerable areas is to use spatial analysis tools available ingeographic information systems (GIS). GIS analysis isdeveloped to examine spatial and temporal patterns and tofind associations between various geographical factors(Mitchell, 1999). Since flooding is a spatial phenomenonand is a consequence of a number of factors (including soiltype, vegetation cover and type, rainfall intensity, rainfallfrequency, etc.) the GIS will allow the user to handle,manage and analyze the spatial data sets to determinewhich factors have what effect and to foresee the resultingconsequences (Melesse and Shih, 2000; Baban and Kantarsingh, 2005). In addition, GIS has the ability tocarryout temporal analysis, which is essential for floodprediction. By storing data on previous floods, soil typesriver channel size etc. it is possible to create a model of peak flow, discharge and runoff. In terms of the impacts of land use/cover on flooding, GIS can be used to detectchange as well as identifying trends, both visually andstatistically, resulting from land use changes and floodedareas (Mamat and Mansor, 1999).The government of Trinidad and Tobago initiated a studyin 1996 to manage flooding in the Caparo River Basin area.The study necessitated the preparation of a watershedmanagement plan for the Caparo River Basin. Furthermore,a GIS-based study was needed to assess the soil erosion,sediment yield, and to locate areas that needed conservationwhen used or expected to be used for agriculture. ARTICLE IN PRESS Fig. 1. Caparo River Basin in the Trinidad and Caribbean context. B. Ramlal, S.M.J. Baban / Journal of Environmental Management 88 (2008) 1131–1140 1132 This paper provides a detailed account of the methodol-ogy used in the GIS component of the study to generateestimates of soil erosion and sediment yields and theidentification of areas needing conservation and the type of conservation appropriate for these areas. Major findingsare discussed. Conclusions and recommendations arepresented. 2. The study area The Caporo River Basin is located in Central Trinidadand is approximately 10km long by 3km wide andrepresents a significant portion of the arable landsavailable. Located in the river basin is the town of Chaguanas, which is the focus of economic activity inCentral Trinidad (Fig. 1).Major land use in this region includes commercialactivities, illegal settlement, slash and burn agriculturalactivities, and quarrying. As such activities in the regionincrease, the natural protective cover of the land is reducedcausing more flooding and erosion. This has led to extensivedamage and loss of valuable raw materials and resources.The flooding problem in the Caparo River Basin is perhapsbest illustrated using the extent of flooding for 1991 (Fig. 2),which covered more than 25% of the watershed. Theproblem is therefore quite significant in this area. To avoidfuture loss, it is necessary to identify the causes of theproblem and attempt to address them urgently. 3. Data collection strategy A thorough analysis of the existing conditions in theriver basin was conducted to provide a basis for furtherdecision making. This involved the collection of historicaldata including flooding, surveys of land use and vegetativecover, rainfall, soil types, geology, elevation, populationdistribution, roads, river network system including riverprofiles and cross-sections, bridges and dams, and delinea-tion of the watersheds and sub-catchments.The data was collected using different surveying andmapping techniques and from various sources such asexisting maps, reports and documentation. A description of how each theme was obtained is provided in Table 1. Fig. 3 shows the elevation, land use/land cover, populationdistribution and soils maps as examples of the base mapsused in the study. The dominant land uses are forests(35%), and agricultural activities (45%). The total popula-tion of the watershed is approximately 37,000 persons. Thewatershed elevation ranges from sea level to 270m highwith more than 78% of the area lower than 100m abovesea level with 94% of the area with slopes less than 30 1 . Thepredominant soil types in each area are Talparo clay(20%), Brasso clay (15%), Ecclesville clay shale (11%), andCunupia fine sandy clay (10%). The average annualrainfall for the upper Caparo watershed area is approxi-mately 2000–2200mm, which decreases to about 1600mmin the coastal area. These data were summarized from the ARTICLE IN PRESS Fig. 2. Flooding extent in the Caparo River Basin 1993 Flood Event. B. Ramlal, S.M.J. Baban / Journal of Environmental Management 88 (2008) 1131–1140 1133 spatial data sets compiled for this project and were used inanalyzing the river system and its catchment. 4. Analysis and results To provide effective mitigation and management mea-sures and to avoid significant erosion and flooding, it isimportant to analyze the soil erosion, sediment transportand river morphological processes of the Caparo River andits catchment (Dion, 2002). A morphological assessment of the Caparo River was undertaken and a sediment transportpredictor was derived. GIS was also used to model floodinglevels in the river basin for different events using 1, 5, 10and 50 year return periods. Based on the findings of theseanalyses, recommendations on the best solutions to use tomitigate and manage the watershed are presented (Stereand Heesterman, 1999). The analyses were divided intothree components: 4.1. The river system The analysis of the specific characteristics of the river isbased on the assumption that the Caparo river system istypically an input–output system (Fig. 4). This implies thatthe output of the system is completely determined by theinputs; which are the basin characteristics and humaninterference. It should be noted that there are independentand dependent variables.The independent variables occur at two levels: at thecatchment level and the level of the reach (Morisawa,1985). At the catchment level, the independent inputs arethe climate and geology of the basin. The climatedetermines the average precipitation and temperature.The geological history is responsible for the rocks thatare present and subject to weathering that leads to soilerosion. The influence of geology and climate is compli-cated by the role of vegetation and the weatheringprocesses. More detailed discussions on the interactionbetween geology, climate, vegetation and weathering maybe found in Morisawa (1985) and Richards (1987). At the river reach level, the independent inputs are the dischargehydrograph, the volume of sediment that has to betransported through the river reach on an annual basis,the characteristics of the sediments whether coarse or finesediments and the slope of the valley. It may be noted thateach of these variables will be different for each river reach.The dependent variables for a river reach are the rivercharacteristics. These may be broken into morphologicaland hydraulic characteristics. These include bed materialcharacteristics, longitudinal slope of the river, channelwidth, number of channels, channel slope, channel depth,water depth, and water velocity in the river (Fig. 4). 4.2. Estimating soil loss and sediment yield Soil erosion is defined here as the amount of soil lossfrom a given slope, usually predicted per unit area basis,and sediment yield is the amount of sediment that passes agiven point on the watershed. Some of the sediment thatleaves a certain slope is deposited; hence, sediment yieldand soil erosion are not the same (Haan et al., 1993).Soil loss can be estimated using the universal soil lossequation (USLE) (Wischmeier and Smith, 1978) whichlends itself to GIS analysis. Improvements to this haveresulted in a modification known as the revised USLE orRUSLE (Renard et al., 1997). The RUSLE is often used toidentify areas that have already suffered damage and thosewhich may be susceptible to damage if not managedproperly.In this case, the revised USLE or RUSLE is moreappropriate since it better estimates the average annual soilloss from runoff for specified cropping and managementsystems (Dion, 2002).The USLE/RUSLE equation is as follows: A ¼ R K L S C P ,where A is the average soil loss per unit area, R the rainfall/runoff factor, K the soil erodibility factor, L the slopelength factor, S the slope steepness factor, C the cover andmanagement factor, and P the supporting conservationpractice.The R-factor is the rainfall and runoff factor. It accountsfor the energy and intensity of rainstorms in RUSLE. Acorrected R -factor in RUSLE reflects adjustments due toponded water on relatively flat slopes associated with heavyrainfall events. In this application, the R -factor is fairlyconstant over the project and it is therefore not taken intoaccount. ARTICLE IN PRESS Table 1Data themes, scale, source, and collection methodsTheme Scale SourceSoils 1:25,000 Existing soil map series sheets33,34, 44 (1975)Elevation 1:12,000 Aerial photographs 1998, fieldsurveysLand use/Landcover1:25,000 Topographic maps, aerialphotographs 1:12,000, field surveysWatershed basins 1:25,000 Derived from 1:25,000 digitalelevation modelRainfall Meteorological office— observations at several stationsclose to study siteRiver network:width, slope,depth, water depth,number of channels1:12000 Aerial photographs, field surveys,river profiles and sections surveyRiver crosssections1:500 Field surveys of river sectionsRoads, bridges,culverts1:12,000 Mapped from aerial photographs,1998.Populationdistribution1:25,000 Census Surveys Conducted by theCentral Statistical OfficeGeology 1:50,000 Existing geology maps (Kugler,1962). B. Ramlal, S.M.J. Baban / Journal of Environmental Management 88 (2008) 1131–1140 1134 The K-factor , also called the soil erodibility factor,quantifies the susceptibility of soil particles to detachmentand movement by water. This factor is used in the USLE tocalculate soil loss by water. It is dependent on various soilparameters like the percentages of silt, sand and organicmaterial and the structure and permeability if the soil.Table 2 shows the relationship between the K -factor andthe various dominant soil types for the project area (Bradyand Weil, 2002).The S-factor , the slope steepness factor is determinedusing the following equations: S ¼ 10 : 8 sin y þ 0 : 03 for sin y o 0 : 09, S ¼ 16 : 8 sin y 0 : 50 for sin y X 0 : 09.These formulae are as proposed by McCool et al. (1987).The L-factor is the RUSLE factor accounting for theeffect of length and steepness of slope on erosion. The ARTICLE IN PRESS Fig. 3. Elevation, land use/land cover, soil and population base data for the study area. B. Ramlal, S.M.J. Baban / Journal of Environmental Management 88 (2008) 1131–1140 1135
