Making Climate Services More Effective

by Sofia Bettencourt*

As demand for climate services diversifies, many institutions have become involved, fragmenting service delivery and rendering it more susceptible to vested interests. Investments are needed in effective and efficient information systems, capable of conveying information to those who need it, where and when they need it. This article examines lessons learned from Africa and the Pacific regions and proposes practical solutions to make climate services more relevant to users.

With the rise of climate-related disasters worldwide, interest in climate services has risen significantly in recent times, expanding beyond basic traditional weather forecasts to long-term trends and early warning systems. National decision-makers, in particular, increasingly ask for climate information to help guide sectoral policies and decrease the risk of disasters. Unfortunately, for most of the developing world, this demand has not yet resulted in effectively linking climate information with decision-making. The problem is institutional, as much as technical.

Decision-makers need climate information, particularly related to:

  • Monitoring and predicting of climate-related hazards so disasters can be averted or mitigated. The key needs in terms of hazards are the expected intensity, frequency and location of events based on both historical statistical analysis and forward looking modelling. Linking hazard analysis to historical exposure and risk modelling is also important to forecast with more accuracy the likely impacts on population and assets.
  • Forecasting climatic conditions for the forthcoming seasons so users can adapt their patterns – for example short-term and seasonal forecasts indicating expected deviation of temperature and rainfall from normal patterns.
  • Forecasting of long-term trends (e.g., changes in the characteristic of extreme events, trends in rainfall and temperature patterns) so they can correctly guide decision-makers in the formulation of key policies and strategies. Here, decision-makers require not only forecasts of climate parameters, but also their expected sectoral and spatial impact.

Traditionally, climate services have focused on short-to medium term weather forecasts (0–14 days) or sub-seasonal (1 month) to seasonal climate forecasts. However, decision-makers increasingly require information on hazards and longer-term climate trends. This requires new types of assessments and more effective and efficient delivery systems, which often transcend the mandates of National Meteorological and Hydrological Services (NMHSs).

Technical constraints remain

Some of the constraints remain technical. Sectoral impact analysis, for example, is an important but often neglected component of climate information, partially due to the uncertainties of current rainfall scenarios and difficulties in downscaling. Yet it remains critically important for strategic planning. When Madagascar adopted its five-year national development plan (2007–2012), it would have been critical for policy makers to know whether its ambitious road infrastructure plan could be supported by the expected flood risk increase in major river basins (Figure 1).

graphic The two figures show the planned expansion of the national road system under Madagascar’s 2007–2012 national development plan. As several key roads are located in flood prone areas, it is important for the achievement of the plan that rehabilitated roads be designed to current and expected flood standards. Source: Madagascar Action Plan 2007–2012.
Source: Madagascar Action Plan 2007-2012

Figure 1 — A potential application of long-term climate trends to development planning

Similarly, as Zambia embarked on an ambitious hydropower and agriculture intensification programme under its Sixth National Development Plan (2011–2015), it was critical to inform policy makers whether future water resources’ balance in the Kafue River basin could support an expansion in power capacity.

Formulation of sectoral impact scenarios is only part of the picture – they also need to be conveyed to decision-makers in a language that they will understand. The 2000 Pacific Island Regional Economic Review: Cities, Seas and Storms – Adapting to Climate Change, addressed this by focusing on four simple questions often asked by policy makers (see Figure 2):

  • What is likely to happen in the future?
  • What are the impacts on key sectors?
  • How much will it cost us?
  • What should we do to adapt?
graphic A: Present status
B: Residual island under a worst case scenario 2100
C: Residual island under worst case scenario and storm surge, 2100
Source: Pacific Island Regional Economic Review: Cities, Seas and Storms – Adapting to Climate Change
Figure 2 — Analysing climate change impacts in the Pacific Islands region

In 1999–2000, the World Bank collaborated with the International Global Change Institute at the University of Waikato and more than 20 agencies to complete an analysis of climate change impacts, economic costs and potential adaptation options in the Pacific Islands region. The study used an integrated climate and sectoral impact model to simulate likely future scenarios of climate change parameters – temperature, rainfall, sea level rise and extreme events – and their expected impact on key sectors, including water resources, agriculture, health, coastal systems and fisheries. The analysis was applied to a representative low island (Tarawa) and a high island (Viti Levu) of the Pacific.

Figure 2 illustrates the potential impact of sea level rise and storm surge on the coast of Bikenibeu Island, South Tarawa. By assessing the opportunity costs of lost assets and economic flows, the study estimated an economic impact on coastal areas of US$ 7.3 million to US$ 12.8 million per year. For all sectors combined, the impact of climate change by 2050 is estimated at 2–4 per cent of Fiji’s Gross Domestic Product (GDP) and 17–34 per cent of Kiribati’s GDP (in 1998 dollars). This study concluded by presenting a range of potential adaptation options to help mitigate the impact. The study has since been used as a basis for the Kiribati Adaptation Program, currently developing its third phase (World Bank, 2000).

Kiribati and Zambia further provide examples of effective mainstreaming. In Kiribati, the Ministry of Finance engaged a strategic planner and public sector expert to help them incorporate adaptation into the 2004–2007 National Development Strategy. In Zambia, the Ministry of Finance and National Planning hired a panel of highly respected experts to help them mainstream the 2011–2015 National Development Plan. In both cases, it was the ministry responsible for economic planning and budgeting that led the initiative, and the embedded experts were highly respected individuals.

By “walking their talk”, the experts were able to interact closely with the economic planners, and thus correctly influence key national planning documents and budgetary allocations. Had this been approached through a stand-alone document, for example sustainable development plans or climate change strategies, the impact would have been much reduced.

In many countries, forecasting abilities are further limited by the fact that historical hydro-meteorological records are incomplete or have not yet been digitized. In Madagascar, the Directorate General of Meteorology addressed this shortfall by collaborating with experts from the University of Cape Town in reconstituting historical records from 21 stations from 1961–2005. This allowed them to estimate past and future climate trends in temperature and rainfall. Using cyclone track modelling they could estimate the expected future trends in cyclone path and intensity. The national team later used this analysis to develop wind-resistant housing standards. Similar background analysis has since been carried out in Mozambique and Sao Tome and Principe.

In 2008, the Directorate General of Meteorology of Madagascar published its analysis of past and future climate change trends. In addition to rainfall and temperature trends, it also analysed future cyclone tracking scenarios. The results were subsequently used to divide Madagascar into four wind-risk zones, and develop new cyclone-safety standards for public buildings. The codes are strictest on the north-eastern shore, where they are set to resist wind speeds of 266 km per hour or 74 m per second. The new legislation was adopted in April 2010.

Institutional constraints and complexity

While technical issues continue to be a constraint, the most important reasons for ineffective climate services tend to be institutional. In both sub-Saharan Africa and the Pacific, colonial powers established meteorological and hydrological networks primarily to serve immediate strategic and military needs. For this reason, stations were often located at airports or major power stations and were not necessarily representative of agro-ecological areas.

After independence, meteorological and hydrological services became intrinsically affiliated with the agencies that their stations directly served – typically departments of transport, communications or aviation (for meteorology services) and departments of water affairs (for hydrological services). During the 1980s, structural adjustments and public expenditure reforms made these agencies prime candidates for budgetary cuts, with the consequence that many meteorological and hydrological stations became degraded or ceased to operate altogether (WMO, 2009). In strategically important sites, such as major mines and infrastructure, private hydro-meteorological stations began to emerge.

This complex institutional setting became further fragmented with the advent of disaster management agencies and, more recently, climate change units, typically run by different traditional “mother” agencies, such as civil defence and environment. Partially as a result of these trends, the WMO 2006 Disaster Risk Reduction (DRR) Country-Level Survey (WMO, 2008) reported only one African country (out of the 28 countries that responded in Africa) and about one-fourth of the Small Island Developing States (14 out of the 19 countries and territories that responded in the Pacific) as having succeeded in combining NMHS under a single agency.

The key difficulty in harmonizing climate information systems lies with budgeting. Government budgets are generally allocated sectorally, making it intrinsically difficult for Ministries of Finance to approve budgets lines covering multiple agencies.

For both donors and recipient agencies, the incentives also favour fragmentation of the system, rather than its consolidation and effective delivery. From the perspective of donors, it is easier to claim a distinct output, such as rehabilitating x-number of meteorological stations or introducing a vulnerability assessment methodology, rather than focusing on an efficient system relying on multi-agency collaboration. Likewise, individual agencies face perverse incentives. Developing their own early warning or monitoring system gives them a strengthened mandate, resources, equipment and opportunities for staff development. By comparison, the incentives toward an effective system are much less clear: Who holds the budget? Who takes the credit? How does one measure an effective system?

This institutional complexity is not helped by the fact that meteorological services and even disaster risk management agencies often lack the mandate to help users adapt to climate variability and prepare for disasters. This is often the mandate of municipal agencies in urban areas, or ministries of agriculture, public works or regional governments in rural areas. Even when early warnings reach the field, funds are often released late, or regional agencies lack the experience or political will to help their communities.

Political complications

A further complication has been the tendency for politicians to use climate information to their advantage. According to an expert from an African meteorological service, how do politicians announce a dry season, when political campaigns have promoted another bumper harvest? Climate scientists are often reluctant to release information that they consider imprecise. Decision-makers, on the other hand, tend to use information that best serves their purposes and ignore information that they consider unfavourable. Uncertain scenarios of rainfall can provide them with just that leverage.

In some regions, critical livestock migration routes in drought-prone areas were reportedly affected by large-scale agriculture investments, promoted based on projections of higher future rainfall by some, but not all, Global Circulation Models. As climate forecasts become increasingly important to the general public, it becomes critical to shield them from political interference, and correctly orient decision-makers and strategic planning documents.

Developing effective climate services

Analyse the key weaknesses in the national systems: The first step is for meteorological services and other involved agencies to analyse the key weaknesses in the national systems: where are the obstacles preventing climate information from reaching end users when and where needed? Figure 3 shows typical components of a national system. Weak links include sectoral impact modelling in long-term climate change forecasts, overly-complex vulnerability assessments that fail to feed into decision-making, early warning systems that are unlinked to funding sources, and the release of timely information to users.

Figure 3 — Typical components of a climate information system

Reinforce institutional links: Over the long term, this second step may be best achieved by linking all key components of the system through coordination mechanisms or programmes, which engage NMHSs, Disaster Risk Management agencies and organizations or institutions responsible for climate change adaptation.

In practice, very few countries have reached this stage. Some countries have combined meteorological with hydrological services (for example, Cook Islands); others have combined NMHSs that run under disaster risk management agencies, such as in Kyrgyzstan. Still others, such as Viet Nam, Tajikistan and Nepal, incorporated NMHSs under environment or natural resource ministries.

Over the short term, many countries have tried to address this constraint by forming multi-agency platforms or committees, or redefining the mandates of the various agencies through legal protocols. In Africa, for example, the WMO 2006 DRR Country-Level Survey found that 30 per cent of the 28 responding countries had established partnership agreements between national meteorological services and national hydrological services (WMO, 2008).

This can be further facilitated by having a single inter-agency committee responsible for disaster risk reduction and for climate change, or by creating a single climate information service platform, as recently done in Zambia.

Make the system more responsive to users’ needs: This third step implies making climate data accessible and understandable to end users (e.g., sectoral users, farmers and community leaders) and insulating it from political interference. To ensure this, NMHSs need to understand the requirements of their users, factor these into the development of their products and services and regularly elicit feedback from them. However, a more challenging task is to ensure that it reaches end users without being “lost in translation”. Apart from the traditional media (radio and television), mobile phone technology has emerged as a promising new tool to promote accountability.

Either through crowdsourcing, freeware or paid services, mobile telephony is increasingly being used to convey weather information and early warnings directly to users, particularly farmers, using short messaging services (SMS). This technology enables NMHSs to send text messages to a large number of users. Sao Tome and Principe, for example, plan to warn artisanal fishers of upcoming storms using SMS.

In turn, users are being encouraged to use SMS to relay information on traditional warning signs – rainfall, flood levels, snow, mudslides, storms gather on lakes, rivers or at sea, etc. Some systems also allow users to reverse SMS charges.

The great advantage of these systems is that once cleared by NMHSs, the warnings are automatically conveyed to end users with minimum political interference. This is particularly the case if NMHSs are able to pass automatic forecasts or warnings emitted by regional centres, such as in the case of tsunamis or cyclones.

Link climate information to financing of risk. Climate risk financing is particularly important in the event of disasters. This is best achieved by establishing contingency funds or insurance schemes that are linked to parametric triggers, rather than relying on declarations of national emergencies, which tend to be more prone to delays and subjective interpretation. Examples of risk financing schemes reliant on parametric triggers include drought insurance schemes in Ethiopia and Malawi, and the Caribbean Catastrophic Risk Insurance Facility (CCRIF).

Linking climate information to risk financing is also important in non-catastrophic events. As inter-annual climate variability increases, there is an increased need for municipalities and local governments to have access to small pools of contingency funds. They can use these funds to adjust their activities in accordance with seasonal forecasts, for example, to repair roads and clean drains likely to be affected by seasonal flooding or to stock food in advance of a drought season. Such systems already exist in Kenya, Ethiopia and Bangladesh, and are under development in the Pacific Islands region.

Suggested ways forward

In sum, the challenges to effective climate services are as much institutional as technical. In a future of increasing climate variability, climate information mandates have become increasingly diffuse, rendering the effective delivery of information and resources all the more complex.

The recent increase in international funding for NMHSs is a welcome trend, but funds need to be channelled into making the system more accessible and understandable rather than contributing to its further fragmentation. To this end, some first steps for NMHSs include:

Facilitating stakeholder meetings to correctly identify the weakest linkages;

Encouraging further institutional consolidation and cooperation among meteorological and hydrological services and agencies responsible for disaster risk management and climate change adaptation;

Promoting accountable systems by conveying climate information directly to end users; and

Ensuring climate information is linked to risk financing mechanisms in ways that can make a difference when and where required.

While these challenges are daunting, many countries have already made important strides towards strenghening climate services. They need to be further encouraged in their efforts.


The author is grateful for the advice provided by Vladimir Tsirkunov, and members of the Global Facility for Disaster Reduction and Recovery Weather and Climate Information and Decision-support Systems


World Bank, 2000: Cities, Seas and Storms: Volume IV – Adapting to Climate Change. (Summary Version), Washington, DC.

WMO, 2008: Capacity Assessment of National Meteorology and Hydrological Services in Support of Disaster Risk Reduction: Analysis of the 2006 WMO Disaster Risk Reduction Country-Level Survey, Geneva.

WMO, 2009: Adaptation to a Variable and Changing Climate: Challenges and Opportunities for NMHSs. (Scientific Lecture, J. Zillman, 11 June 2009, WMO EC-LXI), Geneva.


* Sofia Bettencourt is the Lead Operations Officer, World Bank, based in Lusaka, Zambia. She has managed adaptation and disaster risk management programmes in Africa and the Pacific Islands region since 1999.

Share this page