By switching to dark mode you can reduce the energy consumption of our digital service.

Cape Town Capacity Building

This workshop was held at the University of Cape Town as part of the ACCCA project which is coordinated by UNITAR. For further information on the ACCCA project and the pilot adaptation actions taking place, please visit the project website and search the knowledge base.

The workshop provided a venue for ACCCA project teams to share ideas, interact and learn from the resource persons and from each other about climate adaptation. The objective of the workshop was aimed towards the successful incorporation of vulnerability analysis, climate change data, and risk communications strategies into the ACCCA projects, that will eventually add to the knowledge base on climate change adaptation experiences, methods and tools. The objective of the technical assistance of which this workshop is a part is to support ACCCA project teams in producing and finalising their projects, and to provide a forum for discussion among project teams and between the teams and selected resource persons with expertise in the fields noted above.

Thirteen participants representing ACCCA project teams from Africa (see Annex) as well as a group of relevant resource persons attended and contributed to the workshop in terms of learning about adaptation and exchanging information with other project teams and regional and international experts. Specific outputs of this workshop included an update on the participating project’s progress to date, the identification of immediate steps that would be undertaken by the team upon their return to their respective countries, and a compilation of the specific needs of each country for the successful completion of the ACCCA project.

The workshop was organised by the Climate Systems Analysis Group of the University of Cape Town, with additional support from START, SEA-START, ENDA, SEI and UNITAR.

Workshop Objective and Background

In order to deliver the main outcomes expected from the ACCCA pilot actions, a technical assistance program was developed by the ACCCA support teams (Scientific Advisors; technical backstopping). This included the design and distribution of relevant protocols and products to assist pilot projects in their activities. This included:

  • A protocol for participatory, policy-focused risk communication pilot actions will be tested, evaluated and documented, enabling others to replicate the actions;
  • Risk communication products designed to support adaptation decisions and targeted to decision-makers and vulnerable groups will be developed, tested, and disseminated to engage civil society in the development of adaptation strategies;
  • Recommendations for adaptation and the integration of adaptation into development, environment and resource policies and programs produced and promoted by each pilot action. Recommendations will include the identification of critical knowledge gaps that impede effective adaptation decisions.
  • Plans for new collaborations with stakeholder and scientific organizations that target the identified gaps will be developed and promoted for follow-up action;

Secondary outcomes include:

  • Enhanced capacities of the participating organizations and individuals through ‘learning-by-doing,’ technical assistance and training;
  • Partnerships of stakeholder and scientific organizations will be established for continued cooperation in responding to climate and environmental risks;
  • Networks of institutions in developing country regions will be strengthened for leading south-south capacity building efforts related to the United Nations Framework Convention on Climate Change and other Multilateral Environmental Agreements;


In this context, the overall objective of the workshop was to provide a venue for hands-on work and guidance on the analysis and interpretation of climate data, and on strategies for developing risk communications materials. The working and discussion sessions designed for these days focused on:

  • Analyzing past climatic data and developing envelopes of future climatic conditions relevant to a specific ACCCA pilot action’s needs.
  • Understanding issues of interpretation of climate data, including confidence in climate change data related to geographical and temporal scale, as well as the robustness of climate modeling information.
  • Specifying the links between climatic changes and vulnerable communities that are affected by these changes.
  • Strategies for developing appropriate synthesis materials to support an informed discussion and communication of climate change and vulnerability issues to decision-makers.

Pre-workshop Preparations

The workshop agenda was developed collaboratively developed with the resource persons and discussed/revised prior to the workshop using the online project management tool located at . A final draft was circulated to some members of the team and all other participants prior to the workshop.

The technical support team held a series of preparatory discussions on the agenda and workshop structure, and identified roles and responsibilities for the workshop. The team also reviewed all available documents for ACCCA projects on the online platform at This enabled the support team to have a clear idea of the work conducted to date and to conduct a preliminary assessment of country needs.

As part of the workshop preparations, a resource package was prepared and mailed to each project prior to the workshop. It included background notes and guidance documents of:

Guidance Notes:

  • Identifying Climate Vulnerability ExposureThis guidance sheet provides an approach to identify the key climatic hazards for a specific region and livelihood or economic activity.
  • Risk Communications on Climate Change and VariabilityPreliminary guidance for ACCCA teams. This is a first note for ACCCA teams about to go into the field to communicate risk and the need for adaptation to climate change and variability. A longer document and a training session, together with a more academic background paper will be provided during technical assistance workshops to be held early 2008.

Background Notes:

  • The Climate Change Explorer- ACCCA background note, October 2007.
  • Climate Risk Communication – Background information for Day 3 of ACCCA Cape Town Workshop

Scope of the Report

The workshop report aims to provide a summary of the proceedings (Section III), noting the key points presented by the resource persons (Section IV) available as well as the comments from participating country teams, and the results of the workshop evaluation (Section V).



The workshop was organized by the Climate Systems Analysis Group at the University of Cape Town, with technical support provided by SEI Oxford and, START US (Washington D.C., SEA-START (Bangkok). Funding for the workshop was provided through the generous contribution of the European Commission and the Dutch Climate Change programme (ETC NCAP) and DEFRA UK

All workshop documents, presentations, worksheets, new country documents and materials to review and photos provided by the resources team were posted on the ACCCA project database

Agenda and Participants

The workshop structure was generated ahead of time by the technical support team, and subsequent modifications were made at a pre-workshop meeting held at the University of Cape Town, on Monday 22 October. The Final Agenda in full is available in the accompanying annex as well as the ACCCA project database , along with contact details of resource persons who were present at this workshop.

Report of the Proceedings

Tuesday, 23 October

Welcome and ice breaker

Neil Leary opened the workshop with an ice breaker exercise. The group was asked to team up in pairs and briefly discuss their project approach as well as some of the barriers that have come up during the first months of the programme. The discussions were then briefly reported back to the whole group. Some key points from ice-breaker are summarized here per country:

  • Malawi: floods, problem of food insecurity
  • Zambia: held stakeholder workshop on how people perceive climate change issues; suffer from floods; limited funding is a barrier to work
  • Kenya: convene stakeholder meetings; train health promoters; increase public awareness; establish link between malaria and climate; one of their barriers/challenges is how to communicate technical and scientific information
  • Mali: survey hydrological resources; analyse climate impacts; work on re-vegetation
  • Zambia: work with historical climate data; financial barriers
  • Kenya: water scarcity in drylands; collected community experiences
  • Malawi: community been noticing number of climate variations; been consulting communities about long term trends (e.g. increased frequency of floods); barriers are limited expertise for projecting future climate; notable progress is establishing and improving link between the Red Cross and the Met Service
  • Tanzania: historical memories of communities; they are experiencing no barriers in their project; using an NGO to access villages
  • Ghana: held national forum to increase awareness of impacts of climate on 5 diseases; barriers are accessibility of villages because of extreme weather and the high expectation of villagers
  • Kenya: challenge of dispersed sites that they working in
  • Mali: challenge of missing data and poor station distribution
  • Ghana: been conducting a survey

Session 1: Fundamentals

Bruce Hewitson gave an elaborate presentation on the ins and outs of the use and interpretation of climate data and information. Some of the main messages are briefly summarized here.

Climate data and information

Climate change information is not simply about a change as x or y, or even the mean – climate information may be in the form of an aggregate but it must be remembered that it comes from high resolution, dynamic processes (i.e. rapidly changing weather). There are multiple sources of climate information, there are many attributes to climate data and there are multiple scales (both temporal and spatial) and contexts… and this makes for complexity! One needs to understand the complexities and simplicities associated with climate information; recognise the limits and boundaries; and become comfortable with what uncertainties exist.

In addition some climate information is difficult to get hold of and some information is better than other. There is also sometimes missing meta-data (data about data) and there may be terminology barriers and some information can be presented in difficult formats.

Even though it is important to be aware about these complexities and uncertainties, one should also realize it is not all doom and gloom! One has to accept that there are unknowns but this does not mean that it is impossible to use climate information. It is important, however, to realize that data and information are there to inform choice, not to define choice.

There are some guidelines on how one can operate in such a complex world of incomplete knowledge. Some of those guiding principles are listed below.

  • One needs to understand the context of information to assess the value of the information
  • One must be careful NOT to manipulate data to such an extent that the outcome is meaningless – “one can cut data in many ways but one must not mutilate it”.
  • Climate information is assessed and used through the lens of a specific objective – one needs to make ones own informed/intelligent decision about what data you need and what to do with it
  • One must act responsibly when using data – ask questions of it, don’t simply accept it, “get one’s hands dirty by playing around with the data”.
  • One needs to work with and understand historical data as this is a critical point of departure for exploring future climate
  • Many simply use what is initially measured (e.g. daily rainfall) but it might be appropriate to transform it into something else more relevant to what you are using it for (e.g. dry spell duration). One also needs to consider combinations of different climate attributes e.g. rainfall and wind speed
  • Data is a mix of signal (the information you are interested in) and noise (other variability) – the magnitude of the noise may obscure your signal
  • The historical behaviour of climate is not necessarily what will happen in the future – one can’t simply extend the trend
  • Historical data is useful for setting the context of adaptation, to assess the baseline climate – but first one must decide what baseline is needed, i.e. what attribute is important/relevant to your project and how does one define an appropriate baseline for that.
  • One must also be careful of developing trends from station data because it tends to be messy data.
  • One needs to ask questions like: is the data spatially representative (this depends on topography, such as in Cape Town where rainfall varies dramatically between the mountain areas and the Cape Flats)
  • Observational data, even when gridded, are usually based on point measurements. So the question becomes: what location can I say is realistically represented by the available data. The spatial representation is also dependant on weather, topography and the variable of interest

Some further suggestions (but not a recipe) for handling historical data:

  • Understand the meta-data
  • Quality control – assess the data visually
  • Begin with aggregate measures (e.g. averages)
  • Build to higher orders of variability (if relevant)
  • Examine histograms – look at shape and thresholds (from histogram one can see how often a threshold is surpassed)
  • Do the above for derivative attributes as well
  • Be imaginative, “if you cannot imagine it, it is very unlikely you will discover it”
  • Ask questions: what information can be taken as robust; qualitative versus quantitative conclusions; do the conclusions make physical sense; acknowledge it when something doesn’t make sense

Global/General Circulation Models (GCMs)

  • GCMs are the primary source of information on projecting climate change but they require considerable caution when interpreting the results e.g. one cannot interpret GCM grid cell data at a point, neither can one make place specific statements from GCM results.
  • GCMs are not reality – they are reduced complexity models. GCMs are best at handling large scale processes (spatial and temporal aggregated attributes) – single grid cell data from a GCM is least robust
  • The first step is to validate the model, i.e. can it represent primary processes for the region under study
  • In large areas 2/3s of GCMs do not agree about the direction of change;
  • GCM prognostic variables (variables in the equation, e.g. temperature) are more robust than diagnostic variables (variables as consequences of the model, e.g. precipitation)
  • No one GCM is best or more correct. Therefore one should not throw out a model that disagrees with a range of other models on the variable or region one is interested in – the minority model may in fact be right.
  • One may want to use the model that gets the primary forcing feature right for the climate change projections rather than the one that gets the 20th century ‘right’
  • There are many sources of error in the model – this means that it does some things well but not others

“Don’t tell stakeholders something you can’t justify and defend, rather say that at this point I don’t know but let me find more pieces of information (that provide a context) and then I can give a best guess answer that I can substantiate.”

Accessing data

There are various sources of data available

  • Re-analysis data (possible historical truth)
  • Historical observation data
  • RCM projections
  • Empirically downscaled data

If the ideal data is not available to you then

  • Assess what you can obtain (e.g. global products)
  • Understand it’s limitations
  • Work from a physical based understanding to infer the consequences
Break Out Exercise:Five groups were formed, each representing one sector (Agriculture, Health, Water, Urban management). The groups were asked to think about data needs for carrying out an impact assessment using the following guidelines:
  • Select a geographical location
  • Identify the client
  • What climate data are needed
  • What sources will be used/ how will the data be accessed
  • What are the sources of uncertainty/ what are the limitations of the information/ what are the expectations for the outcome

Downscaling – Regional Climate Models and Statistical Downscaling

There are two ways of downscaling – Regional Climate Models and empirical/statistical downscaling

Regional Climate Models (RCMs) involve using a high resolution, limited area, dynamical model, forced at the boundaries by a GCM. It produces gridded cell data (usually 25-60km squared grid cell)

  • Regional Climate Models are based on fundamental physics and usually only use 1 or 2 driving GCMs, but may have ensembles of more GCMs.
  • RCMs are vulnerable to incorrect representations of local processes. Like GCMs a single grid cell value is of low robustness, hence one should not interpret a single grid cell but rather the surrounding.
  • RCMs have significant logistical requirements (e.g. large data files, arcane formats) and they introduce their own bias and error
  • RCMs are in some ways an interim tool, until GCMs can run at high resolution

Empirical/statistical downscaling produces comparable outputs to RCMs but are easier to implement and manage. Empirical/statistical downscaling is based on the fact that local scale climate is largely a response to large scale forcing and is good at capturing first order responses

There are two conceptual approaches empirical/statistical downscaling:

  • Model stochastic element produced by a weather generator and this is conditioned by large scale forcing e.g. SDSM tool
  • Represent cross scale relationships between large scale and local climate and add the stochastic residual (as used in ACCCA)

There are some warnings that go with downscaling. First of all: downscaling can go badly. Downscaling does not seek to reproduce the real world observation for observation but rather to generate a realistic time evolution. Also it is important to remember that downscaling does not correct for error in the GCM

Important messages
  • Characterise baseline observational climate as best possible
  • Characterise process change to inform understanding
  • Use many models – look at that ranges
  • Downscale if possible
  • Remember that uncertainty can arise from many sources. However, this does not mean that nothing can be stated robustly.

Session 2: Refining user needs

During the second session Tom Downing introduced a framework for identifying and refining the climate information needs in terms of what kind of climate indicators and thresholds are important. The framework can be considered as a simplified version of the livelihood sensitivity matrix. Figure 1 gives an example how this could work for the agricultural sector.

Image:Igure1 framework.jpg

Figure 1: Framework for identifying climate information needs

Starting from socio-economic vulnerability, one can identify the impacts and adaptation strategies and from there explore the relevant climate indicators and thresholds. Considering crop growth in the agricultural sector, important climate indicators are for instance seasonal rainfall and probability of drought over 3 consecutive years. Equally important would be to look into indicators such as: the drivers of vulnerability; policy; stakeholders; and value chain. However, this was not the focus of this session. The project groups were then asked to discuss and think about what kind of climate information would be needed for their specific projects, using the following three steps;

  1. Identify and describe the vulnerable group
  2. How is the climate impacting this group
  3. What are the important climate variables/indicators and thresholds associated with these climate impacts

Day 1 wrap up session

Reactions to ACCCA Management Platform and Progress Indicators:

Participants were asked provide their reactions to the ACCCA management platform. Half the group indicated apprehensiveness and confusion about the tool; nearly half indicated acceptance of the tool but with some reservations; and 1 person indicated enthusiastic support. Most participants will need time for hands-on use and exploration of the platform to gain a better sense of how to use the platform, the potential benefits and the cost in terms of time.

Progress indicators will require some negotiation with advisors to come to common understanding the interpretation of the different targets within the context of individual projects and to agree on appropriate indicators. It was suggested that the technical assistance teams also make entries in a monthly journal to provide their perceptions of progress on targets as a complement to those of the pilot team leaders.

The message about the need to use and keep current the project tasks on the management platform needs to be communicated to pilot team leaders who are not present at the Cape Town workshop.

Reactions to session on principles for approaching climate knowledge:

Many of the participants indicated that they found the presentation difficult to follow, much of the language was too technical and that the presentation ‘went over our heads’. There was a sense of being overwhelmed by the complexities and uncertainties and not knowing what the principles imply for execution of the pilots or what to do next. One person observed that ACCCA is not supposed to be a research project but the morning’s presentation seems to imply that they are expected to do more research and analysis than is feasible within the limited budgets and timeline; a desire was expressed to simply have the needed data prepared and delivered to the teams by CSAG.

Others indicated that they enjoyed the presentation and found it useful; that it was good to see that there are African climate scientists upon whom they could rely. A lesson from the day is that teams need to develop understanding of the different climate analysis and modeling tools in order to have confidence in their outputs and to make informed judgments about what are the climate data requirements of their projects and what approaches are needed/appropriate.

Other comments on the first day:

  • Shared lodging arrangements were a surprise to some participants; would have appreciated advance notification that they would be sharing accommodations.
  • More variety and less pasta/lasagna for lunch.
  • Can return flights be confirmed?

Wednesday, 24 October

Session 3: Exploring the future

The objectives of this session were to get insight into data for one’s region and explore what one might do with this data in a simple spreadsheet. This was run as a very hands on session with 2 participants to a computer and a number of facilitators and demonstrators from the CSAG group in support.

Two sets of data were provided:

  1. multiple GCM data
    • timeseries of monthly means from 11 models
    • long term climatology (long term average for each month)
    • control climate (1960 – 1999)
    • future projection (2046 – 2065)
    • data are spatially smoothed (grid cells for a single location reflective of the regional response)
  1. sample downscaled data
  • point location downscaled from GCMs
  • past and future data
  • daily time-series

With this data people explored the seasonal cycle, looked at averages across the models, calculated the standard deviation of all the models to look at the spread, looked at the difference between future and past climates, etc.

Session 4: From interpretation to action – developing key messages for adaptation

The data and where they come from – University of Cape Town

Chris Jack presented the UCT web-based data portal. The web portal contains data sets for current and future climate in Africa. A brief overview of the main navigation tools and functionalities was given.

Currently they are developing a new and updated web portal which will feature more information about the different climate models as well as provide the possibility of downloading data in a format that is compatible with the Climate Change Explorer tool.

Downscaled climate data

Bruce Hewitson presented an overview of and briefly commented on the main downscaled climate parameters that are currently available. Data and information about other parameters, which are not yet included in the list, can also be provided if needed.

During the discussion the use of weather generators for exploring possible future climates was brought up. Bruce commented that weather generators have some serious disadvantages, which makes them not very suitable for making future climate scenarios.

One of the participants asked whether information about anomalies will be made available in future. Bruce answer was that it can be created.

As a final point of the discussion, Bruce observed that there are two kinds of comparisons made in climate change science. Firstly, model results for a certain period are compared to observed data for that same period in order to get a measure for bias and error in the model. Secondly, model results for both current and future climate are compared to assess changes. It is important to notice that model results for future climate should not be compared with observed data from the past.

The Climate Change Explorer tool

Stewart Collis and Fernanda Zermoglio of AWhere Inc, presented and illustrated the use of the Climate Change Explorer tool. The Climate Change Explorer is a new tool that seeks to make climate data analysis easier for a range of practitioners, researchers and decision-makers. It addresses this by packaging data access routines, guidance and customised analytical and visualization procedures that provide users a sound analytical foundation from which to explore climate variables relevant to their adaptation decisions. Groups got the opportunity to explore some of the functionalities of the Climate Change Explorer.

Challenges in disseminating scientific information to local communities

Gilbert Ouma, a project member for the ACCCA Kenya Drylands project, presented some experiences with disseminating scientific information about climate change to local communities in Kenya. They started with questions how people receive information about climate and weather related information. Next, they tried to find out how and if people are using that information. It quickly appeared that people often did have access to climate and weather related information through newspapers, TV and radio, however, nobody really relied on that information. Most people rather turn to traditional sources of weather forecasts.

In an attempt to change these attitudes, scientists started discussing with local communities the backgrounds and science behind climate and weather related information. At first, the community was overblown by the information but in the end, people started to understand that climate information is based on something and started to accept the information provided by the scientist. Hence, by working closely together with the communities, the group of scientists gradually gained confidence and they developed a common ground for understanding.

Mukufute Mukelabai presented another experience from Zambia in which they went to remote communities to provide climate and weather related information. They assembled large community meetings. Initially, however, people would not tend to believe such information. Hence, the strategy taken was to start from the knowledge of the local people and find ways to complement it with scientific knowledge. This collaboration between ‘traditional’ knowledge and scientific knowledge appears to have been key to the success of the programme.

Another example, presented by Francis, came from Ghana where women are responsible for fetching water from often remote places. From a medical point of view, the incidence of diseases among women and children is much higher and there seems to be a close relationship with the tasks women are typically carrying out. When trying to motivate men to take up their responsibilities in this, the programme encountered severe resistance from the men (‘who are you to make judgments about our culture?’).

Thursday, 26 October

Session 5: Introduction to risk communication

During session 5 the participants were split up into two groups to prepare a presentation on climate risk as set out on a prescribed summary provided. Two speakers were then elected from each group to make a ten-minute presentation to the other group. The information and tools provided were very limited so that space was created for speedy identification of good communicators. The exercise was also deliberately constrained by time in order to increase the pressure on participants to make quick decisions.

It appeared from this role play exercise how difficult it is to produce a cohesive risk communication, especially if time is limited. Moreover, good communicators and not so good communicators became quickly apparent.

The role play exercise was followed by a more theoretical session on the fundamentals and principles of risk communication, including a discussion on how to define risk in general and climate risk in particular.

Session 6: Communication skills

The communication skills session looked to describe different aspects of communications skills and how they can be employed usefully and effectively. Lawrence Flint talked about the importance of identifying good communicators and the opportunities and threats associated with different communication strategies. The session also emphasised the complimentarity and interlinkages between the physical and social sciences in work on climate vulnerability and adaptation.

Session 7: Conceptualizing the social dimensions of adaptation

During this session discussions took place amongst the project participants about what adaptation is really about. It tried to add some further dimensions to the adaptation debate by including the right to development and issues of equity as a the fundamental basis for adaptation.

Session 8: Practical Exercise

Building on the experience from the first role play exercise during session 5 and utilizing some of the ideas and skills discussed during the other sessions, the participants were again split up into two groups to develop a good risk communication strategy. This time, more time was given to refine the approach and to create a good presentation.

The workshop closed with an interactive evaluation the results from which are available in the report’s Annex.


Add your project

Exchange your climate change adaptation projects and lessons learned with the global community.