Storm Surge Mapping for Montego Bay, Jamaica
Executive Summary

Organization of American States
General Secretariat
Unit for Sustainable Development and Environment

Report prepared by Smith Warner International Ltd., Kingston Jamaica

This executive summary is extracted from the full report entitled Storm Surge Mapping for Montego Bay, Jamaica.

Executive Summary

Introduction

The General Secretariat of the Organization of American States (OAS), in collaboration with the USAID, has sponsored the Caribbean Disaster Mitigation Project (CDMP). As part of this CDMP, a storm hazard pilot project was developed for Montego Bay. The intention of this project was to fine-tune the surge computation and analysis procedures, so that these could be used with confidence in other countries of the Caribbean. In addition, present regulatory and risk assessment practices were evaluated in the light of the predicted storm surge findings.

At the start of the project, a computer model was used to give storm surge predictions for Montego Bay. Specifically, the output from the model included:

• Maps showing the Maximum Envelope of Water (MEOW) generated surge.
• Storm surge predictions for 10,25, 50 and 100 year return periods.

The methodology for the computation of storm surge, and the manipulation and interpretation of this data, has subsequently been subjected to a detailed review process. This has resulted in the development of a well-documented approach to storm surge computation, which has been aided by the input of Professors Robert Sorensen and Mark Johnson (of Lehigh University and University of Central Florida respectively) and which has been accepted by the engineering regulatory body in Jamaica. It is believed that the methodology proposed may now be applied with confidence throughout the Caribbean region.

This report forms the final document within the Montego Bay Storm Hazard Project. Three main components have been included. The first is the recommendation for a policy framework, for the use of storm hazard information in Jamaica. In developing this policy framework, an assessment has been made of the current practice regarding the use of hazard information by planners, engineers, and the insurance industry.

The second part of the report includes a set of guidelines for the computation, interpretation, use and presentation of storm hazard information. As part of these guidelines, some basic definitions for storm surge have been made.

The third section of the report gives a description of a case study that was undertaken for Doctor’s Cave in Montego Bay. Within this section, the storm surge models used, have been described.

Finally, the report includes a description of the entire project development, including papers presented at the October 1997 Technical Workshop. These have been presented as Appendix 1.

Current Practices and Regulatory Approach

In investigating the existing policy framework for storm hazard computation and analysis, a review was made of current practices and the regulatory approach to risk. Under this category, design practices advocated by the Jamaica Institution of Engineers were examined. These revealed well defined design criteria for: wind loading on buildings (using a 1 in 50 year return period event); earthquake loading using the Uniform Building Code approach; and flood levels in drainage works (with return periods ranging from 1 in 2 years for culverts, up to 1 in 50 years for major flood control works).

The procedures employed by the Office of Disaster Preparedness and Emergency Management (ODPEM) were also examined. It was determined that ODPEM’s primary goals, of public education and emergency management, could be greatly enhanced by the acquisition of appropriately scaled mapping. This mapping should ideally be at a scale detailed enough to show infrastructure. Storm surge elevations for varying return periods should then be superimposed onto this base mapping.

Insurance industry practices and needs were then examined. It was assessed that in order to improve the risk quality of its insurance portfolios, it was essential for the industry to develop an in-depth knowledge of the contributors to this risk. Information that was considered to be important for this assessment included:

• Hazard mapping of threatened areas as a function of risk (such mapping should be done for hurricane winds, surges (i.e. flooding), earthquakes, etc.);
• Vulnerability of critical infrastructure, with the implementation of hazard mitigation measures and adherence to building code recommendations; and
• Adoption of sustainable criteria upon which to base recommendations for development.

The present regulations in Jamaica regarding set-back of infrastructure from the mean water line are presented in terms of foreshore slope and are not related to actual risk occurrence. The review carried out of the Town Planning Department regulations and practices clearly indicates the urgent need for revision of these setback regulations, to provide a clear linkage between setback limits and acceptable risk. One prerequisite for this type of approach is the adoption of an acceptable return period that will provide adequate protection against an acceptable level of risk. Recognizing this need, the suggestion has been put forward that the 1 in 25 year return period be adopted by the regulatory planning agencies. Given the adoption of this level of risk, the computed surge elevation may then be drawn onto a suitably scaled base map. From this superposition, information may then be obtained regarding the vulnerability of infrastructure within the coastal zone. In addition, it has been recommended that the Town Planning Department maintain a mandatory minimum setback from the HWM, of 15 metres. This minimum setback is intended to incorporate safety considerations for both cliff-type shorelines and also for erosion considerations for beach-type shorelines.

Finally, as no appropriate code presently exists for the estimation of wave loading in the landward storm surge area, guidelines for the computation of wave heights at the shoreline have also been put forward.

Storm Surge Definitions

The actual components of storm surge have been defined as being the sum of three primary components:

1. Meteorological effects caused by the low atmospheric pressure at the center of a storm;
2. Wind effects caused by the wind stresses on the water surface; and
3. Wave effects caused by waves breaking in the surf zone (known as wave set-up).

These three components contribute to a static water level, which therefore does not include wave run-up, or the crest elevation of the wave above the mean water level. In addition, in the computation of storm surge, it is important to consider tidal variations and also long term sea level rise. For this latter component, a UNEP recommended rate of sea level rise of 5 mm/year is proposed.

Analysis and Presentation of Data

Once storm surge values have been computed for the long-term database, these must then be subjected to a suitable statistical analysis. From this analysis, it will be possible to estimate storm surge corresponding to a variety of return periods. Two statistical methods have been presented for handling hurricane/storm surge data. The first method is an Annual Maximum Method, which requires the computation of a maximum value of storm surge for each year of record. The second method is a Peak Over Threshold Method, which requires the identification of storms passing within a user-specified distance from the location of interest. Results from each method were found to be very similar for the more extreme events (e.g. over 1 in 25 years), but the Peak Over Threshold method resulted in higher value predictions for more frequent events.

The visual presentation of this data has been described in the report, and the following steps are proposed:

1. First, a base map must be prepared on which all generated data can be superimposed. This mapping is typically obtained either through the generation of ortho-rectified images, or through the use of geo-referenced Survey Department maps. Depending on the scale of mapping used, data pertaining to infrastructure location should be included.
2. The base map should contain topographic information, which may be in the form of land contours or spot elevations, or both.
3. The minimum requirements of appropriate mapping for storm surge delineation should be at a minimum scale of 1:5,000 with contour intervals of 1.0 metre for the first ten metres in elevation above datum.
4. The inclusion of land use data is strongly recommended. Data pertaining to land use zones may be obtained (in the case of Jamaica) from the Town Planning Department’s 1993 Development Order.
5. The map should also contain enumeration data. This may be obtained, in the case of Jamaica, from the STATIN 1991 Census data, with the coastal zone data extracted.

Recommendations for Use/Application of Storm Surge Data

It is recommended that a rational approach to the regulation of setbacks should be implemented. These setbacks need to be based on actual statistical risk, in order to provide a consistent means of protection along an island’s shoreline. The legal mechanism for this approach would be through Development Orders, which should include the demarcation of a flood line. The main drawback to the implementation of this approach pertains primarily to the requirement that storm surge data must first be generated for the entire coastal zone (in a lateral sense), and appropriate mapping (minimum scale of 1:5000) must be available.

Engineering considerations within this zone include the following:

• Flood proofing of structures including the placement of critical infrastructure above the predicted inundation level. Alternatively, flood proofing may be achieved by designing the ground floor of buildings to be dispensable during major surge events.
• Buildings that may be threatened by wave action within an inundation zone should be designed to withstand the forces resulting from the expected wave height at these locations. The Shore Protection Manual (USACE, 1984) contains the procedures for computing wave forces on structures.

Finally, for the proper application of emergency management procedures, it is essential for these agencies to have access to mapping showing maximum inundation limits corresponding to varying categories of storms (in accordance with the Saffir Simpson scale). This mapping may then be used to identify the most vulnerable infrastructure, and as well, to estimate the number of people who may be threatened by flooding, for a particular category of tropical storm or hurricane.

Results of Case Study

A case study was carried out for the Montego Bay area, and in particular at Doctors Cave. Computation of storm surge was carried out using both the TAOS model with an annual maxima approach, as well as the generic model, with a peak-over-threshold approach. These analyses gave the following results:

1. For the TAOS/Annual Maxima approach:

Montego Bay Storm Surge (m) using 2-Parameter Weibull Distribution (Doctors Cave)

2. For the Alternative/Peak-Over-Threshold approach:

Doctor's Cave Storm Surge (m) using Weibull Distribution k=1.4

Using the alternative approach, storm surge values for varying return periods were also computed for the Montego Bay Airport, the Montego River outfall and the Montego Freeport. The 1 in 25 year return period (MLE) surge computed at these locations were 1.6 m, 1.7 m and 2.5 m respectively, indicating a range of almost 1.0 metre depending on the exact location.

In the presentation of the computed data, a base map was prepared from 1:15,000 scale 1991 colour photography of the area. Ortho-rectification of these images was achieved by following a novel process, whereby the available data was used to produce a detailed base map image at a relatively low cost. The ortho-rectification process involved the following steps.

• Aerial photograph contact prints (at 1:15,000 scale) were scanned at 600 dpi resolution to create digital aerial imagery. Un-published Survey Department 1:2,500 maps were then scanned and geo-referenced.
• Plan control for scaling and orientation within the national grid was extracted from the 1:2,500 maps and vector elevation coverage. The digital aerial imagery was geo-referenced.
• Land elevation contours were digitized (heads up) from the raster maps, along with the elevation points shown on the ground, to create a digital vector coverage of the study area. This would facilitate the creation of a DEM for terrain relief correction.
• The digital aerial image was then ortho-rectified using pixel-by-pixel single frame techniques available in ERDAS IMAGINE 8.3 software. The individual frames were then re-sampled and a mosaic image created. The image base map and vector contour coverage was suitable for importing into ESRI ARCVIEW 3.1 software for GIS analysis.

The computed storm surge results for the case study have been mapped within a GIS, which has been used to evaluate infrastructure and population areas at risk. The coastal zone of the Montego Bay area is planned for resort, resort/residential, transportation, conservation, commercial/office, institutional, light industry, open space, public assembly, public buildings, agriculture and residential uses. Mapping of the 1 in 25 year return period surge (see Figure ES-1) indicates that of a total land area of 1400 hectares within this zone, approximately 480 hectares (or 34%) will be impacted by storm surge. In addition, it was found that land areas associated with conservation, resort and transportation (airport) were the most seriously affected.

The population within the coastal zone was obtained from the 1991 Census data, to be 16, 626. Significant growth of this number is not expected, since the area is already well developed and there is little provision for new housing other than through the redevelopment of existing housing lands. Mapping of the storm surge shows that the 1 in 25 year event will not have major impacts on this population. By contrast, the 1 in 50 year event is shown to have significantly more impact and would require mitigation planning by ODPEM. The physical problems likely to be experienced by these communities during the close passage of hurricanes are attributable to low level flooding and public health risks associated with contaminated water.

Finally, the mapping of storm surge showed that some critical infrastructure would be threatened during the passage of a storm (see Figure ES-2). These included: the new Divisional Police Station; the new main NWC sewage pump station; and the Sangster International Airport. Special emergency management plans will need to be formulated to ensure that as soon as the elevated waters subside, these facilities may be returned to active service. In addition, the banks of the North and South Gullies, and the Montego River, are likely to be overtopped, as elevated water levels at sea will affect the flood regimes in these watercourses.

Figure ES-1. Inundation Area for the 1 in 25 Year Hurricane Surge

Figure ES-2. Impacts of Storm Surge on Population/Critical Infrastructure

This executive summary is extracted from the full report entitled Storm Surge Mapping for Montego Bay, Jamaica.