**1. Introduction**

532 Risk Management – Current Issues and Challenges

http://new.colyirr.com.au/

\_update/en/index.htm

[61] Sanders F, Burt M. Subcritical Contraction for Improved Open-Channel Flow Measurement Accuracy with an Upward-Looking ADVM (with Charles M. Burt and

[63] Health issues gain traction at UN Climate Conference [Internet] WHO; 2011. Available from:http://www.who.int/globalchange/mediacentre/events/2011/durban\_conference

[64] Protecting health from climate change - World Health Day 2008 [Internet] WHO; 2008. Available from: http://www.who.int/world-health-day/toolkit/report\_web.pdf. [65] Disaster Risk Management for Health Fact Sheet [Internet] WHO; 2011. Available from:

[66] Confalonieri U, Menne B, Akhtar R, Ebi K, Hauengue M, Kovats R, Revich B, Woodward A. Human health. Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, M.L. Parry, O.F. Canziani, J.P. Palutikof, P.J. van der Linden and C.E.

[67] Ebi K, Woodruff R, Hildebrand A, Corvalan C. Climate change-related health impacts

[68] Casimiro E, Calheiros J, Duarte F, Kovats, S. National assessment of human health effects of climate change in Portugal: Approach and key findings. Environmental

http://www.wmo.int/pages/themes/climate/risk\_management\_overview.php

http://portal.iri.columbia.edu/portal/server.pt?open=512&objID=1094&mode=2 [71] The Social Dimensions of Climate Change, Discussion draft [Internet] 2011. Available from: http://www.who.int/globalchange/mediacentre/events/2011/social-dimensions-of-

[72] IPCC. Special Report on Managing the Risks of Extreme Events and Disasters to

[73] Samazzi F.H.M. Framework for climate services in developing countries. Climate

Brett F. Sanders), Journal of Irrigation and Drainage Engineering; 2010.

http://www.who.int/hac/events/drm\_fact\_climate\_risk\_management.pdf

Hanson, Eds.,Cambridge University Press, Cambridge, UK; 2007: 391-431.

[69] Climate Risk Management Overview [Internet] WMO. Available from:

in the Hindu Kush-Himalayas. Ecohealth;2007(4): 264–270.

Health Perspectives; 2006 (114): 1950–1956.

[70] IRI Resources tools [Internet]. Available from:

Advance Climate Change Adaptation; 2010.

climate-change.pdf

Research; 2011 (47): 145-150.

[62] Coleambally Irrigation [Internet]. Australia 2009. Available from:

The *Observatorio Latinoamericano de Eventos Extraordinarios* (OLE2)*,* or *Latin American Observatory*, is a regional collaborative network that, in addition to the existing infrastructure in each country, aims to ultimately increase the efficiency of the decisionmaking processes, especially in terms of getting more accurate environmental information and exchanging experiences on data, methodologies and scientific products, all of which are done with a standardized methodology and a Web-sharing service.

Present partners of this collaboration are national weather services, universities and research institutes of Central America, the Andean countries and from Southeast South America (see Figure 1). Nowadays, the Centro de Modelado Científico (CMC) of Universidad del Zulia, the founder of the initiative, regionally coordinates the OLE2. The coordinating role involves, among other tasks, the suggestion of (a) methodologies for the provision of scientifically based tools and products, (b) mechanisms for the successful use of the environmental information in the policy making process at the different levels (e.g. national and province governments, private sector), (c) facilitation of continuous and effective communication of the different partners and (d) the interchange of experiences and products (by means of the Observatory's products web interface1, wiki page2, forum3 and email list4). For more information see references [1] and [2].

<sup>4</sup> ole2@cmc.org.ve

© 2012 Muñoz et al., licensee InTech. This is an open access chapter distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. © 2012 Muñoz et al., licensee InTech. This is a paper distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

<sup>1</sup> http://ole2.org

<sup>2</sup> http://mediawiki.cmc.org.ve

<sup>3</sup> http://laft.cmc.luz.edu.ve:8080/foro\_ole2

The goal of the OLE2 is to monitor and forecast key environmental variables and develop accurate products based on different scientific tools in order to help decision makers improve risk management and set up efficient early warning systems. The Observatory provides several model outputs for meteorological, seasonal, and hydrological forecasts, 5 day high-resolution oceanographic prediction for the eastern Pacific, droughts, fire and flood indices, ecosystem dynamics (like duckweed/algae occurrence in Lake Maracaibo), and climate and health applications (e.g., regarding malaria), among others.

Risk Management at the Latin American Observatory 535

**Figure 1.** The Observatorio Latinoamericano's partners (as January 2012) are mainly national weather services, universities and research and development institutes. (The red line links the countries

Among the different efforts, risk assessment and management are some of the most important tasks targeted by the Latin American Observatory partners. Even when a general framework is used by the OLE2 on this field, the different institutions develop tailored applications to take advantage of the local experience and mechanisms already in existence in their countries. However, the general framework closely follows the ideas discussed by

Risk management is from a wide perspective, a policy system that, among its components, permits stakeholders in several levels, to identify, analyze and quantify the probability of human, material and ecosystem losses, as well as to provide metrics and technical criteria for prevention, mitigation, reduction or transfer of risk [4]. A key idea to be still implemented

participating in the AndesGrid iniciative [3]. See main text).

Mora [4].

The Observatory is organized in interconnected working groups (WGs) or axes: namely, meteorology; climatology; hydrology; air quality; climate and health; and climate change and variability. Each WG has its own methodologies and products, but all of them exchange information and results with other axes and countries if necessary. It is also regionally divided in three branches: the Observatorio Andino, Observatorio Centroamericano and Observatorio del Sudeste de Sudamérica (Andean, Central American and the South Eastern South America Observatories). They operate using the infrastructure already in existence in the different regions by consolidating the collaboration networks through the recognition of common needs and the provision of interdependent solutions. The final products, be it a seasonal forecast or an early warning system for the last-mile user, is built between different partners by sharing their experiences and expertise on different fields. This has created a strong identification between the participating institutions and the Observatory's products. This fact and the continuous need of collaboration to provide regionally successful decisionmaking tools on time, have proven a natural way to guarantee the sustainability of the initiative, originated in 2007 in Venezuela as a national observatory.

Advances and difficulties are shared and discussed through e-mail lists and videoconferences. The purpose of the videoconferences is to discuss the ongoing projects, products, and new methodologies, and to provide special assistance on models and several other technical issues.

The Observatory has also provided a Grid Technology Infrastructure since 2010, known as AndesGrid [3] to help institutions with less computational power to increase their capabilities using other institutions' available computer resources. The initiative enables the partners to share not only model outputs and, but also their local observations (e.g. rain gauge and temperatures) in "real time". For additional details, see [3].

Of the OLE2's main achievements in this period, the most important is the enhancement of the institutions' human resources by means of continuous technical support for forecast, modeling, verification, risk assessment and management and many other issues, all of which is done in their native language. The project has also succeeded in standardizing forecasts, data formats, and methodologies, providing common models, tools, and procedures that are used on a daily basis in all the involved institutions. A key tool has been the wiki, which contains all of the needed steps for each task, which by itself is another of the most important elements for ensuring the long-term continuity of the OLE2.

other technical issues.

The goal of the OLE2 is to monitor and forecast key environmental variables and develop accurate products based on different scientific tools in order to help decision makers improve risk management and set up efficient early warning systems. The Observatory provides several model outputs for meteorological, seasonal, and hydrological forecasts, 5 day high-resolution oceanographic prediction for the eastern Pacific, droughts, fire and flood indices, ecosystem dynamics (like duckweed/algae occurrence in Lake Maracaibo),

The Observatory is organized in interconnected working groups (WGs) or axes: namely, meteorology; climatology; hydrology; air quality; climate and health; and climate change and variability. Each WG has its own methodologies and products, but all of them exchange information and results with other axes and countries if necessary. It is also regionally divided in three branches: the Observatorio Andino, Observatorio Centroamericano and Observatorio del Sudeste de Sudamérica (Andean, Central American and the South Eastern South America Observatories). They operate using the infrastructure already in existence in the different regions by consolidating the collaboration networks through the recognition of common needs and the provision of interdependent solutions. The final products, be it a seasonal forecast or an early warning system for the last-mile user, is built between different partners by sharing their experiences and expertise on different fields. This has created a strong identification between the participating institutions and the Observatory's products. This fact and the continuous need of collaboration to provide regionally successful decisionmaking tools on time, have proven a natural way to guarantee the sustainability of the

Advances and difficulties are shared and discussed through e-mail lists and videoconferences. The purpose of the videoconferences is to discuss the ongoing projects, products, and new methodologies, and to provide special assistance on models and several

The Observatory has also provided a Grid Technology Infrastructure since 2010, known as AndesGrid [3] to help institutions with less computational power to increase their capabilities using other institutions' available computer resources. The initiative enables the partners to share not only model outputs and, but also their local observations (e.g. rain

Of the OLE2's main achievements in this period, the most important is the enhancement of the institutions' human resources by means of continuous technical support for forecast, modeling, verification, risk assessment and management and many other issues, all of which is done in their native language. The project has also succeeded in standardizing forecasts, data formats, and methodologies, providing common models, tools, and procedures that are used on a daily basis in all the involved institutions. A key tool has been the wiki, which contains all of the needed steps for each task, which by itself is another of

and climate and health applications (e.g., regarding malaria), among others.

initiative, originated in 2007 in Venezuela as a national observatory.

gauge and temperatures) in "real time". For additional details, see [3].

the most important elements for ensuring the long-term continuity of the OLE2.

**Figure 1.** The Observatorio Latinoamericano's partners (as January 2012) are mainly national weather services, universities and research and development institutes. (The red line links the countries participating in the AndesGrid iniciative [3]. See main text).

Among the different efforts, risk assessment and management are some of the most important tasks targeted by the Latin American Observatory partners. Even when a general framework is used by the OLE2 on this field, the different institutions develop tailored applications to take advantage of the local experience and mechanisms already in existence in their countries. However, the general framework closely follows the ideas discussed by Mora [4].

Risk management is from a wide perspective, a policy system that, among its components, permits stakeholders in several levels, to identify, analyze and quantify the probability of human, material and ecosystem losses, as well as to provide metrics and technical criteria for prevention, mitigation, reduction or transfer of risk [4]. A key idea to be still implemented

in most parts of the world is that risk management needs to be understood and addressed as an investment, not as a cost. [4] It is clear that nowadays the establishment of efficient risk reduction plans does not possess a real priority on the agenda of decision-makers and stakeholders, except *a posteriori*. This is especially true in most parts of Latin America, as is exposed in some case studies included in this chapter. Moreover, in general, there are no consolidated cultures incorporating risk management throughout decision-making processes for public and private investment and planning (see [4] and references therein). Also, private sector involvement in this process presents several unique challenges to be overcome [5].

Risk Management at the Latin American Observatory 537

**2. Risk definition and management** 

outline the main ideas for future reference.

internal geodynamics (seismicity and volcanism)

,

identification, analysis and quantification of risk

financial, social and environmental protection

communicate their products to the end-mile users.

always be carefully considered. Moreover, *V* and *H* are not independent.

terms of their origin, namely

weather extreme events)

according to the following relationship:

causes

In this and the next section the main definitions related to risk management, hazards and vulnerabilities are discussed in order to provide the foundations chosen by the Observatory in these matters. We don't pretend to address all the details in this complex field, but to

Different agents play a role in the occurrence of natural hazards, which derive from the damaging potential or a combination of them. These agents are frequently [4,6] classified in

hydro-meteorology: both global and local processes (hurricanes, floods/droughts,

Following Mora [4], natural hazards (*H*) can be formally defined in terms of the probability that an event becomes so intense (*a*) within time and space frames that it produces significant damage (*d*). The intensity and damage definitions depend on the region and timescale under consideration. Vulnerability (*V*) can be associated with the probability that, according to the intensity of the natural event, damage might occur as a function of the degrees of exposure and fragility of the elements involved. Risk (*R*) therefore is the combined probability (convolution, \*) that a hazard might cause significant damage,

() ( ) \* ()

The probabilities are not, in general, constant, so the spatial and temporal variability must

Risk management involves not only the quantification of the probability of loss via equation (1) and the derived secondary effects, but also the generation and execution of policies for [4]

formulation and application of measures of prevention, reduction and mitigation of

 execution of related protocols (e.g. preparedness, response, rehabilitation, reconstruction). Efforts to foster these processes may include the establishment of adequate policies and protocols to incorporate stakeholders into risk management, improvement of the climate/geodynamic information available, identification of the different sources of vulnerability and their evolution in time, definition of "acceptable" risk levels and the establishment of early warning and monitor systems, and the adequate ways to

*<sup>p</sup> R da p H da p V da* (1)

*h d h d*

external geodynamics (landslides, intensive erosion, torrential debris flows).

Since 1975 in Latin American and the Caribbean (LAC), the costs of losses attributable to major natural hazards have been estimated at around US\$300 billion, with more than 280,000 human deaths and affecting more than 160 million people [6-8]. Table 1 references the major natural disasters in LAC in the last three decades.


**Table 1.** Major natural disasters in LAC between 1983 and 2011 [6-8].

Consequently, it has been an urgent need for the Latin American Observatory's partners to address the establishment of effective policies for risk assessment and management in their respective countries. To this end, an implementation plan was suggested by CMC and this chapter presents briefly the adopted definitions and general ideas (sections 2-4), as well as case studies for different managerial sectors in Latin America (section 5) in order to discuss in some detail, present risk assessment/management policies and methodologies. Finally, we discuss the benefits of sharing experiences among different countries to address common problems in geographically complex regions, as well as how to achieve success in countries where the scarcity of economic and trained human resources imposes serious limitations on the effectiveness of risk management systems.

### **2. Risk definition and management**

536 Risk Management – Current Issues and Challenges

overcome [5].

2010, 2011

in most parts of the world is that risk management needs to be understood and addressed as an investment, not as a cost. [4] It is clear that nowadays the establishment of efficient risk reduction plans does not possess a real priority on the agenda of decision-makers and stakeholders, except *a posteriori*. This is especially true in most parts of Latin America, as is exposed in some case studies included in this chapter. Moreover, in general, there are no consolidated cultures incorporating risk management throughout decision-making processes for public and private investment and planning (see [4] and references therein). Also, private sector involvement in this process presents several unique challenges to be

Since 1975 in Latin American and the Caribbean (LAC), the costs of losses attributable to major natural hazards have been estimated at around US\$300 billion, with more than 280,000 human deaths and affecting more than 160 million people [6-8]. Table 1 references

1983 Floods Argentina, Bolivia, Brazil, Peru

1999 Floods and landslides Venezuela (Vargas State)

2009 Drought (strong El Niño event) Venezuela, Colombia and Ecuador

Consequently, it has been an urgent need for the Latin American Observatory's partners to address the establishment of effective policies for risk assessment and management in their respective countries. To this end, an implementation plan was suggested by CMC and this chapter presents briefly the adopted definitions and general ideas (sections 2-4), as well as case studies for different managerial sectors in Latin America (section 5) in order to discuss in some detail, present risk assessment/management policies and methodologies. Finally, we discuss the benefits of sharing experiences among different countries to address common problems in geographically complex regions, as well as how to achieve success in countries where the scarcity of economic and trained human resources imposes serious limitations on

Floods (strong La Niña event) Panama, Colombia and Venezuela

Central America and the Caribbean

Central America and the Caribbean

the major natural disasters in LAC in the last three decades.

Year Hazard/Event Countries

1985 Earthquake Mexico

hurricanes George and Mitch

2010 Earthquake Chile

**Table 1.** Major natural disasters in LAC between 1983 and 2011 [6-8].

1998 Floods, landslides associated to

2005 Most active hurricane season recorded in history [8]

the effectiveness of risk management systems.

1983 Earthquake Chile, Colombia

In this and the next section the main definitions related to risk management, hazards and vulnerabilities are discussed in order to provide the foundations chosen by the Observatory in these matters. We don't pretend to address all the details in this complex field, but to outline the main ideas for future reference.

Different agents play a role in the occurrence of natural hazards, which derive from the damaging potential or a combination of them. These agents are frequently [4,6] classified in terms of their origin, namely


Following Mora [4], natural hazards (*H*) can be formally defined in terms of the probability that an event becomes so intense (*a*) within time and space frames that it produces significant damage (*d*). The intensity and damage definitions depend on the region and timescale under consideration. Vulnerability (*V*) can be associated with the probability that, according to the intensity of the natural event, damage might occur as a function of the degrees of exposure and fragility of the elements involved. Risk (*R*) therefore is the combined probability (convolution, \*) that a hazard might cause significant damage, according to the following relationship:

$$\int\_{h,d} p(R)da = \int\_{h} p(H)da^\* \int\_{d} p(V)da \tag{1}$$

The probabilities are not, in general, constant, so the spatial and temporal variability must always be carefully considered. Moreover, *V* and *H* are not independent.

Risk management involves not only the quantification of the probability of loss via equation (1) and the derived secondary effects, but also the generation and execution of policies for [4]


Efforts to foster these processes may include the establishment of adequate policies and protocols to incorporate stakeholders into risk management, improvement of the climate/geodynamic information available, identification of the different sources of vulnerability and their evolution in time, definition of "acceptable" risk levels and the establishment of early warning and monitor systems, and the adequate ways to communicate their products to the end-mile users.

Risk management policies foster and integrate strategically balanced processes capable of providing decision-making tools in a synergistic, decentralized and participative way. In the OLE2 this is a central philosophy, which also enables the treatment of risk in trans-boundary zones (e.g. the Observatory enabled the continuous exchange of climate services between the National Weather Services and decision makers during the 2010 floods that affected Panama, Colombia and Venezuela in similar ways).

Risk Management at the Latin American Observatory 539

*RD D D RM* (2)

( ) *B DD I RM RM RM* (3)

(4)

**Figure 2.** Benefit/cost (B/C) ratio defined as a function of damage expected without risk management (D), reduced by applying risk management (DRM) and the investment involved (IRM). (After Mora [4])

It is possible to use risk management as a tool for providing an idea of the economic value of potential damages caused by hazards and the actions undertaken to reduce vulnerabilities. Figure 2 sketches an accumulating damage curve *D* without risk management. If risk management policies are used, this same curve is reduced to a more optimistic distribution,

The associated investment (*IRM*) for such a reduction depends on both the hazard and vulnerability involved (which, again, are not independent). On the other hand, the Benefit/Cost (*B/C*) ratio allows for the finding of optimal levels of investment. It's possible to

> / 1 *RM RM*

Having real distributions of the different quantities it might be possible to select optimal points (e.g. the saddle point named *B/COPT* in Figure 2) to limit profitable investments. See

The traditional approach to computing risk probabilities using equation (1) involves the consideration of static or, in the best-case scenario, long-term mean vulnerability maps of

*D D B C I*

**3. Benefits of risk management** 

*DRM.* The reduction of damage due to risk management is then

show [4] that the net benefit of risk management is

**4. A Multi-scale approach for risk management** 

and the Benefit/Cost ratio can be defined as

[4] and [6] for details.

An important aspect of risk management is related to operative platforms that enable decision makers and users to interact and to obtain information useful to assess vulnerabilities, the impact of natural hazards and the risk probabilities for a certain region. This infrastructure is aimed at providing this kind of information *ex ante*, but it can also be used during or after an emergency. It usually includes


The implementation of these platforms in LAC is more advanced in a few countries (see section 5.1 for a Central American example). The Observatory's partners are taking advantage of the experience developed by other members of the regional initiative to improve their own risk management infrastructure.

Finally, but no less importantly, the Observatory is now committed to following the International Research Institute on Climate and Society's (IRI) Four Pillars for Climate Risk Management [31], namely


**Figure 2.** Benefit/cost (B/C) ratio defined as a function of damage expected without risk management (D), reduced by applying risk management (DRM) and the investment involved (IRM). (After Mora [4])

#### **3. Benefits of risk management**

538 Risk Management – Current Issues and Challenges

time (see section 4) damage assessment

reclamation contracts) information on past events

Management [31], namely

using that information.

favorable climate conditions.

Panama, Colombia and Venezuela in similar ways).

used during or after an emergency. It usually includes

identification of priority areas of intervention

Risk management policies foster and integrate strategically balanced processes capable of providing decision-making tools in a synergistic, decentralized and participative way. In the OLE2 this is a central philosophy, which also enables the treatment of risk in trans-boundary zones (e.g. the Observatory enabled the continuous exchange of climate services between the National Weather Services and decision makers during the 2010 floods that affected

An important aspect of risk management is related to operative platforms that enable decision makers and users to interact and to obtain information useful to assess vulnerabilities, the impact of natural hazards and the risk probabilities for a certain region. This infrastructure is aimed at providing this kind of information *ex ante*, but it can also be

macro- and micro-zoning of hazards and vulnerability and its probable evolution in

administrative and legal information (e.g. design, operation, lease, transfer and

The implementation of these platforms in LAC is more advanced in a few countries (see section 5.1 for a Central American example). The Observatory's partners are taking advantage of the experience developed by other members of the regional initiative to

Finally, but no less importantly, the Observatory is now committed to following the International Research Institute on Climate and Society's (IRI) Four Pillars for Climate Risk

1. **Identify vulnerabilities and potential opportunities** due to climate variability/change

2. **Quantify uncertainties in "climate information"** in order to reduce uncertainties in

3. **Identify technologies and practices** that optimize results in normal or favorable years

4. **Identify interventions, institutional arrangements and best practices** that reduce exposure to climate vulnerabilities and enable the opportunistic exploitation of

as well as technologies and practices that reduce vulnerabilities.

protocols for vulnerability reduction (hazards are not, in general, reducible)

 instruments for financial strategy and budgetary allocations identification of "risk bearers" and their responsibilities

areas of further investments to reduce future disasters.

improve their own risk management infrastructure.

for a given water, agriculture, or health system.

It is possible to use risk management as a tool for providing an idea of the economic value of potential damages caused by hazards and the actions undertaken to reduce vulnerabilities. Figure 2 sketches an accumulating damage curve *D* without risk management. If risk management policies are used, this same curve is reduced to a more optimistic distribution, *DRM.* The reduction of damage due to risk management is then

$$RD = D - D\_{RM} \tag{2}$$

The associated investment (*IRM*) for such a reduction depends on both the hazard and vulnerability involved (which, again, are not independent). On the other hand, the Benefit/Cost (*B/C*) ratio allows for the finding of optimal levels of investment. It's possible to show [4] that the net benefit of risk management is

$$B\_{RM} = (D - D\_{RM}) - I\_{RM} \tag{3}$$

and the Benefit/Cost ratio can be defined as

$$B \,/\, \text{C} = \frac{D - D\_{\text{RM}}}{I\_{\text{RM}}} - 1 \tag{4}$$

Having real distributions of the different quantities it might be possible to select optimal points (e.g. the saddle point named *B/COPT* in Figure 2) to limit profitable investments. See [4] and [6] for details.

#### **4. A Multi-scale approach for risk management**

The traditional approach to computing risk probabilities using equation (1) involves the consideration of static or, in the best-case scenario, long-term mean vulnerability maps of

the associated hazards. Nonetheless, the probability density function associated with hazards in equation (1) is frequently computed from time series that may contain useful information at different time scales. Given the fact that a decision maker might be interested in a certain time scale (e.g. on the order of its own managerial period), a temporal decomposition of the original time series used to identify the hazard probability may be extremely helpful.

Risk Management at the Latin American Observatory 541

In this section a few state-of-the-art examples of risk assessment and management in several countries in Latin America are briefly described. Signatory and participant institutions of the Latin American Observatory operate over these structures, providing tailored

**5.1. Seasonal Climate outlooks applications for food security decision-making in** 

In Central America, seasonal climate outlooks turn into risk scenarios used by food-related sectors to help support their decisions and prevent food insecurity. This is a coordinated

The Central American Climate Outlook Forum (CA-COF), coordinated by the Comité Regional de Recursos Hidráulicos (CRRH-SICA)5 (Regional Water Resources Committee), has consolidated a process to issue three seasonal outlooks per year, bringing together the capacities of all seven weather services in the region. The CA-COF takes advantage of international and global sources of information, which it analyzes along with its own sources and historical data to produce Climate Outlooks for the Central American region.

To facilitate the use of climate risk information for decision-making with the aim of reducing food insecurity risks, CRRH-SICA has fostered a mechanism over the past few years that turns the Outlook into risk scenarios for those sectors related to nutrition and food security, particularly agriculture, fisheries, potable water and public health, as well as cross-cutting areas like risk management and emergency response. These scenarios are used when deciding preventive measures to mitigate the impact of climate variability on food security.

In a joint effort between the specialized agencies of the Central American Integration System (SICA) and the Regional Food Security Program (PRESANCA II) funded by the European Union, immediately after an Outlook is issued, a group of regional experts is convened to enhance the Outlook with information from the different sectors mentioned above, turning the product into sectoral climate risk scenarios that could guide early warnings of actual and

The working group, composed of CA-COF members and the experts, uses these scenarios to identify sector-specific preventive measures. The food security risk scenarios and the suggested measures are circulated among government entities and other organizations involved in addressing food and nutritional security through their own networks to ensure

5 CRRH is the technical Secretariat of the Central America Integration System, responsible for the coordination of activities related to weather forecast, climate, water resources and climate change assessments with Belize, Costa Rica,

methodologies based on what has been explained in previous sections.

effort with specialized entities of the Central American Integration System.

Over the last decade the Forum has issued 38 regional climate outlooks.

all this information reaches the most appropriate decision makers.

**5. Cases of study** 

**Central America** 

potential hazards to food security.

El Salvador, Guatemala, Honduras, Nicaragua and Panamá.

For example, Figure 3 presents such time-scale decomposition for Southeast South America's precipitation. The lower panels sketch the non-linear long-term trend (a proxy for climate change), and decadal and interannual variability time series [9]. The rationale is that in considering the interaction of the three climate signals (the explained variance being a way to ponder the specific weight of each one of them) it is possible to provide better climate services to decision-makers, that are more adequate for their scale of interest, e.g. next 10 years.

**Figure 3.** Decomposition for Southeast South America's precipitation (boxed region) for three different time scales: long-term trend (left panel), decadal variability (central panel) and interannual variability (right panel), with their respective explained variance (colors in the map and values in the bottom corner of each time series). Grid boxes in white are not statistically significant at a p-level=0.95. (After Goddard *et al.* [9])

As mentioned before, hazards and vulnerabilities are not independent. The latter encompasses exposure to hazards, sensitivity to these hazards and adaptive capacity. All of them, in fact, evolve in time. Therefore, a similar multi-scale time decomposition must also be done to correctly assess the related vulnerabilities. This provides a more realistic risk probability distribution, and thus better assessments and policies can be developed.
