Climate Services – Transitioning from CLIPS to GFCS

By Govindarajalu Srinivasan¹, Roberta Boscolo², Tanja Cegnar³, Caio Coelho⁴, Tobias Fuchs⁵, André Kamga⁶, John Kennedy⁷, Vlidimir Kryjov⁸, Barbara Tapia⁹ and Ryuji Yamada¹⁰

In 1995, the World Meteorological Congress established the Climate Information and Prediction Services (CLIPS) project “to provide the best possible climate information, including expectations of future conditions, to improve economic and social decisions that will reduce risks and improve economic vitality as well as quality of life.” Since, CLIPS has strived to increase climate knowledge, improve operational climate prediction capabilities, and develop the capacities of National Meteorological and Hydrological Services (NMHSs) to deliver climate information to meet the needs of stakeholders.

CLIPS was instrumental in the development of the concept of Regional Climate Centres (RCCs) and their formal establishment on a global scale. It also played a key role in the development of the Regional Climate Outlook Forums (RCOFs) that now serve as platforms for generating consensus-based seasonal climate outlooks. These initiatives, together with the CLIPS training workshops, have helped build capacities for climate services.

The present Global Framework for Climate Services (GFCS) with a vision “to enable society to better manage the risks and opportunities arising from climate variability and change, through the development and incorporation of science-based climate information and prediction into planning, policy and practice” carries forward and builds on the solid foundation laid by CLIPS. So, while acknowledging the significant contributions made by CLIPS, the 2011 World Meteorological Congress endorsed closing CLIPS by 2015 and assimilating its activities into the emerging GFCS Climate Services Information System (CSIS), including relevant linkages with the GFCS User Interface Platform (UIP) component.

A WMO Technical Conference (TECO) Climate Services – Building on CLIPS legacy, held from 30 June to 2 July 2014 at Heidelberg, Germany, discussed the challenges of transitioning the project into the relevant components of the GFCS. The Conference was held in conjunction with the WMO Commission on Climatology (CCl) and the Joint Scientific Committee of World Climate Research Programme (WCRP). The conclusions of the discussions are summarized below.

Climate data management

Good quality data is the bedrock of climate services. Recognizing this is an important prerequisite for providing effective climate services. Presentations at the Conference covered Climate Data Management Systems (CDMS) challenges and emerging technologies to meet these challenges. Discussions brought out the need to facilitate access to quality-controlled and interoperable datasets with user-friendly interfaces for deriving relevant climate information products and for supporting new strategies being evolved by WMO Members. The need for optimizing the use of available climate data is also recognized within the WCRP Grand Challenges. Increasing demand for location-specific climate information requires credible climate data at finer resolution.

Observed data needs to be well organized and accessible in electronic CDMS. NMHSs worldwide are now implementing such systems. A dedicated CCl Expert Team has been working on CDMS for storage and management of meteorological datasets. CCI has also been making efforts to ensure standards and build capacities of NMHSs in the area of data storage. Analysis of a survey conducted in July 2013 showed that most NMHSs have computerized climate databases, with about 75% of them using Relational Database Management Systems. However, the CDMS were not fully or properly implemented in about 40% of NMHSs. Eighteen percent of the NMHSs surveyed still used outdated systems and 50% of them required manual intervention to generate routine reports. Dedicated efforts are needed to put in place robust climate databases at national levels to generate quality climatological information. The Bureau of Meteorology, Australia, has deployed CliDE (Climate Database for Environment) in 14 Pacific countries as a part of an ongoing capacity development effort.  

Learning from the experience of CLIPS implementation, the quality and homogeneity of climate data must be ensured in order to provide the best possible climate information. As quality control features were not built into data management systems in the past, it is necessary to ensure that this is a standard feature of present-day CDMS.¹¹ Systematic biases in such long-series datasets must be recognized and corrected through homogeneity testing. Changes at station locations, changeover to automated systems and urban heat island effects are some factors that influence homogeneity. Data rescue is a first step for developing regionally and nationally relevant long-term and high-quality climate datasets.

Metadata must also be properly managed as they can help to detect and interpret data errors. Robust approaches for dealing with various aspects of homogeneity are required and NMHSs should be provided with guidance and tools to archive metadata and ensure homogeneity of their climate datasets. The consistency of data quality is pivotal for the monitoring of climate trends.

Strategies to integrate climate data from a wide variety of sources, including satellite and other remote sensing platforms, while maintaining homogeneity standards are important for providing climate data at high resolution. This will be a real challenge, which some NMHS have started to address while in the same instances providing data at one kilometre spatial resolution on an hourly basis.  

Climate monitoring and assessment

Climate monitoring is crucial for assessing the state of the climate system, climate variability and change. A comprehensive suite of climate monitoring products will improve national capacities for providing high-quality science-based climate services that remain consistent at different levels. There is a need for synergy between forecasting and climate monitoring in order to provide confident climate information and services. A CCl Task Team aims to provide a list of climate monitoring products – for which guides, standard software tools and training will be developed – that can be consistently generated by most NMHSs.

Progress has been made in addressing the challenges of creating long-term climate datasets. For example, the International Surface Temperature Initiative (ISTI)¹² will publish verifiable and unrestricted temperature datasets at both monthly and higher – daily and sub-daily – resolutions. These data will support a suite of monitoring products from hourly to century timescales and from location specific to the global mean. The first full version of the ISTI databank, featuring monthly data from approximately 32 000 stations in over 200 countries, was recently released. Enhanced contribution of additional datasets by NMHSs will be very helpful for the success of this project.

Satellite based generation of climate data has recently demonstrated considerable progress. For example, the European Organization for the Exploitation of Meteorological Satellites (EUMETSAT) Satellite Application Facility on Climate Monitoring has developed a sustained capability for comprehensive generation and provision of climate data Records for the Global Climate Observing System (GCOS) Essential Climate Variables.

Climate prediction, projection and delivery mechanisms

CLIPS implemented modern techniques for issuing seasonal forecasts using operational multi-model ensemble systems. Currently, efforts are also being made to integrate empirical and dynamical systems to improve seasonal forecasts in projects such as the Euro-Brazilian Initiative for improving South American seasonal forecasts.

Global long-range climate predictions on seasonal scales are based on simulations by atmospheric and coupled atmosphere-ocean general circulation models that require considerable information technology infrastructure as well as highly skilled human resources. The WMO Global Producing Centres of long-range forecasts (GPCs) form the core of the GFCS CSIS. Currently, there are 12 formally designated GPCs under the purview of WMO. The information they provide is pivotal for minimizing risks in decision-making. There is urgent need to link seasonal forecasts to application models as illustrated in the Brazilian example (below) about dengue outbreak potential.  

Dengue probabilistic outlook for Brazil, June 2014: an early warning model framework driven by real-time seasonal climate forecasts. Category boundaries defined as 100 cases per 100 000 inhabitants and 300 cases per 100 000 inhabitants. The greater the colour saturation, the more certain the forecast of a particular outcome (Source: Lowe et al., 2014¹⁴)

Sub-seasonal and seasonal predictions are challenging tasks as predictability at such time scales is relatively low, compared to short-range forecasts, because of significant uncertainties arising from the various climate system components and their complex interactions. Significant progress in understanding uncertainties and producing useful probabilistic forecasts has been achieved through ensembles of climate predictions from independent models. This has led to a growing recognition that international exchange of climate prediction information is essential for improving predictability and of the need for creation of mechanisms to facilitate such exchange.  The WMO Lead Centre for Long Range-Forecast Multi-Model Ensemble and the WMO Lead Centre for Standard Verification of Long-Range Forecasts were established to enhance and facilitate the international exchange for seasonal prediction data. They enable NMHSs to generate standardized seasonal forecasts with adequate verification statistics for ensuring confidence in these products.

The Global Seasonal Climate Update (GSCU), a new product being developed by WMO, aims to monitor the state of the seasonal climate and predict its future evolution on the global scale, based on products from GPCs, in order to assist the NMHSs as well as RCCs and RCOFs. To be in time for RCOFs, GSCU production will need a six-month integration. The GSCU will be one of most comprehensive forecast products based on an ensemble of 13 state-of-the-art operational models. After the completion of a process of broad-based scoping and a pilot phase, the GSCU will be ready for operational production.

Decision-makers need climate information tailored for specific time frames. Climate projections are important for developing long-term adaptation strategies. It is important to clearly understand the stakeholder and end-use context in order to provide climate data relevant to specific needs. The value added by downscaling is an ongoing question, particularly in the context of climate change projections that cannot be verified. Nevertheless, regional climate models are important tools that need to be carefully examined and appropriately interpreted for various decision-making contexts. Efforts like the WCRP Coordinated Regional Downscaling Experiment provide the larger ensembles of simulations necessary for exploring and understanding uncertainties in regional climate change projections.

User interface for climate adaptation and risk management

Climate can be viewed as a risk as well as a resource. An effective interface with users that can deliver climate information that is clear, precise and relevant is an important aspect of climate services. Working with stakeholders involves social issues, science and continued communication–such as online discussions, e-mails, videoconferences, social-networks and text messages on cellular networks.

There is currently a movement towards “impact-based forecasting” that produces “impact variables” relating directly to stakeholder activities. Impact-based forecasting suites that link climate forecasts to impact models can be operationalized by scaling up successful demonstration projects. The EUPORIAS project aims to develop and deliver a reliable impact prediction system through five semi-operational prototypes using seasonal to decadal climate information.

As mentioned at the outset, one of the major components of CLIPS was the RCOFs. Climate scientists, policy-makers and other stakeholders join together during RCOFs to understand potential impacts of seasonal climate outlooks on key sectors. RCOFs and RCCs represent successful building blocks for developing climate services. The infrastructure built by CLIPS at global and regional level needs to be further strengthened with the recent advances in science and further modernization of infrastructure in order to improve climate service quality and communication.

For optimal decision-making, research outcomes and climate forecasts information must be interpreted and translated into useful and usable information, thus RCOFs include a thematic session focused on decision-making processes. RCOFs strive to tailor information and bridge the gap between the scientific community, operational systems and practical needs. Indeed, climatologists need to understand a wide spectrum of needs. Training of NMHS staff is required and infrastructure and networking or partnerships with other stakeholders are key requirements for a responsive and effective participation of countries in future RCOF expansion programmes.

The GFCS is expected to provide a vehicle for taking the progress made at the global and regional levels to the national levels. National Climate Outlook Forums (NCOFs), anchored by the NMHSs, provide regular, systematic and sustained interaction platforms. They bring together national stakeholders from agriculture and other sectors sensitive to climate, national partners agencies, universities, non-governmental organizations, research institutions, and local and government committees on climate change and development to define needs and to understand and make best use of climate information.

Regional and national consultation meetings held by GFCS have identified the following 10 priorities for operationalizing climate services at national level:

• provide a strong institutional anchorage for climate services;
• meet the demand for tailored climate service provision in priority climate-sensitive sectors;
• improve communication a for wider distribution of climate services;
• diversify communication channels and use innovative broadcast channels;
• modernize and increase the density of the national hydrometeorological observing network;
• improve collaborative climate research, towards more salient end-user driven climate research outputs;
• build the capacity of the NHMS and other technical services to jointly elaborate salient climate products and services by building on multi-disciplinary knowledge and expertise;
• develop and strengthen the capacity of stakeholders to engage in the production and use of climate services;
• define sustainable frameworks for climate services at the national level; and
• engage all national stakeholders involved in the production, interpretation, communication and use of climate services in a national dialogue around climate service provision in order to identify country needs and chart a course for the provision of user-tailored climate services.  

Capacity development issues

CCl is developing a toolkit to help NMHSs provide climate services while minimizing additional workload. The toolkit, a software suite, will include climate data management and analysis and prediction packages, accompanied by training modules. It will support the generation and use of climate information and prediction products for specific needs. It will facilitate production, communication and application of climate information products and help promote standardization and implementation of best practices in all countries. It will offer an opportunity for least developed countries to use state-of-the-art methodologies for climate monitoring and prediction. The toolkit is expected to be available in multiple languages, easy to use, portable for offline use or requiring minimum Internet bandwidth, and written in free software or freeware.

There is a need for tools that work across a range of space and time scales, incorporating impact models of differing complexity and uncertainty information, from sub-seasonal to centennial scales. Uncertainty needs to be accounted for at all stages. Global models, downscaling, impacts models, all involve significant uncertainty. Decision-support must also be integrated into tool development and an iterative process is needed to ensure that the user’s needs and capabilities are known. The iterative process will potentially need to address capacity building and user feedback.

Future research and operationalization

The Conference agreed to address the following topical issues of direct relevance to climate risk management and adaptation, particularly to the GFCS:

1. Enhance efforts to implement CDMSs and expand data rescue and quality control. The consistency between different climate components, information; namely monitoring and forecasting products should be strengthened, especially in terms of data used, base period and methods.
2. RCCs and RCOFs need to be further expanded and strengthened by introducing more systematic and objective procedures, including verification methods to enhance confidence. In view of the increasing demand for sub-seasonal information, the progress on seasonal to sub-seasonal prediction should be leveraged for operational services.
3. Implement systems to channel recent research efforts enhancing the use of predictions at sub-seasonal and seasonal–to-interannual and decadal scales through operational systems put in place at global, regional and national levels for providing high quality climate services to stakeholders.
4. Future development of the GSCU will need to improve post-processing move from quarterly to monthly issues, focus on RCOF needs and, consequently, extend the forecasted period.
5. Collaboration between CCl and WCRP in areas of sub-seasonal prediction is needed for improving skill of seasonal predictions, understanding and prediction of extremes, downscaling climate information and exploring teleconnections to understand climate variability.
6. The Climate services toolkit is seen as a set of software products designed to support the generation of climate information and prediction products for user needs. A workshop to discuss design specifics – languages, trainer and certification and identification of potential candidate applications would be beneficial.
7. Capacity development (both for individuals and institutions) and recognized as fundamental for good quality climate services.
8. In the process of developing products, stakeholders across the whole value chain need to be involved. This will make climate information valuable for decision-making by being specific and tailored.
9. There is a need for more interdisciplinary work to better address specific needs in various sectors.
10. Access to relevant datasets should be improved for monitoring, downscaling and tailoring for effective climate services relevant to users.

Over the last 20 years, CLIPS has supported the provision of climate services. CCl has also fosters the taking of high quality observations, managing electronic climate data archives and coordinating the use of those data in climate monitoring and assessment. These activities have helped CCl take a lead role in the development of the GFCS CSIS.

Related Links

TECO2014 Materials

Acknowledgments

This article is based on the session reports coordinated by G. Srinivasan with inputs from the session rapporteurs B. Tapia, A. Kamga, R. Yamada, C. Coelho, J. Kennedy, T. Fuchs, T. Cegnar, V. Kryjov and R. Boscolo; and session chairs N. Plummer, J.-P. Ceron, S. Sensoy, M. Semawi, R. Graham, A. Tait, B. Nyenzi, F. Semazzi, F. Doblas-Reyes, A. Martis, A. Busalacchi and T. Peterson.  The article was reviewed by the CCl Management Group and WMO Secretariat.

¹ Regional Integrated Multi-Hazard Early Warning System for Africa and Asia (RIMES)
² World Climate Research Programme (WCRP)
³ Meteorological Office, Slovenian Environment Agency
⁴ Centro de Previsão de Tempo e Estudos Climáticos (CPTEC), Instituto Nacional de Pesquisas Espaciais (INPE), Brazil
⁵ German Meteorological Service/Deutscher Wetterdienst, Germany
⁶ Climate and Environment Department, African Centre for Meteorological Applications for Development (ACMAD)
⁷ Climate Research and Monitoring, Met Office
⁸ Federal Service for Hydrometeorology and Environmental Monitoring, Russian Federation
⁹ Chilean Meteorological Service
¹⁰ Tokyo Climate Centre, Japan
¹¹ WMO, 2014: Climate Data Management System Specifications, WMO-No. 1131, pp. 166.
¹² Thorne, Peter W., and Coauthors, 2011: Guiding the Creation of A Comprehensive Surface Temperature Resource for Twenty-First-Century Climate Science. Bull. Amer. Meteor. Soc., 92, ES40–ES47. doi: http://dx.doi.org/10.1175/2011BAMS3124.1
¹³ EUROBRISA http://eurobrisa.cptec.inpe.br/
¹⁴ Lowe, R., Barcellos, C., Coelho, C. A. S., Bailey, T. C., Coelho, E. G., Jupp, T., Graham, R., Massa Ramalho, W., Sá Carvalho, M., Stephenson, D. B., Rodó, X, (2014). Dengue outlook for the World Cup in Brazil: an early warning model framework driven by real-time seasonal climate forecasts. Lancet Infectious Diseases, 14:619-26. doi: 10.1016/S1473-3099(14)70781-9.