Transmenu powered by JoomlArt.com - Mambo Joomla Professional Templates Club

In terms of using the probabilistic projections to assess current and near term (next 5-10 years) vulnerabilities, impacts, risks and adaptation decisions, users should be aware of what is and is not available from the UKCP09 probabilistic projections and what this means for how they can and should not be used. The first 30-year time period for which probabilistic projections are provided is the 2020s (average projected climate and change for the period 2010-2039). In addition, the baseline period is 1961-1990. There is, therefore, no information within the probabilistic projections covering the period between this baseline period to the 2020s.

As such, the UKCP09 probabilistic projections cannot be used as the sole basis for impacts, vulnerability and adaptation assessments for near-term and current climate.

Alternative Approaches

Find out more

• See FAQ: How do I deal with changes in the UK since the 1961–1990 baseline?

Further reading

• Probabilistic forecasts of near-term climate change based on a resampling ensemble technique. Räisänen, J. and Ruokolainen, L. 2006. Tellus 58A, 461–472 (2006).

• Probabilistic forecasts of near-term climate change: sensitivity to adjustment of simulated variability and choice of baseline period. Ruokolainen, L. and Räisänen, J. Tellus, 59A, 309–320. (2007).

The value added by the UKCP09 probabilistic projections comes from the quantification of the nature and extent of uncertainties associated with the provided projections and its implications in terms of the "quality" of the resulting decisions. Users should, therefore, take full advantage of this enhancement within their decision-making processes and not limit their considerations to a single value.

Users should recognise that there is no magic or special significance associated with the central estimate or value identified as relatively the most likely within the Probability Density Function.

The realisation of this added value is particularly evident as decision makers move away from attempting to identify and implement what they believe may be the optimum adaptation strategy to an approach based on resilient and adaptive strategies reflective of the projected climate, including its uncertainty. Both the literature and experience support that this is an essential move when dealing with the uncertainties that are (and will continue to be) part and parcel of future climate change projections.

Further information

The traditional adaptation assessment approach focuses on identifying an optimum adaptation strategy that addressing the climate risks associated with a single projection or limited sampling of projections (e.g. that provided by a deterministic projection or a user-defined “best” estimate such as the mode, median or central estimate within a probability distribution function). Strategies based on a limited consideration of the possible climate by nature are quite brittle (subject to failure).  They lead to those using a limited approach accepting a higher risk than necessary.  Strategies based on this limited search for an optimial solution could prove far from optimal if the future turnes out differently than that suggested by the adopted projections.

User should embrace the value added by the quantified uncertainties provided by the probabilistic projections. This means not automatically reverting to considering a single projection or only a limited user-defined sample around a central measure (mode, median or central estimate). Rather users should consider the sensitivity of decisions and strategies across the provided probabilistic projections in a manner that includes a judgement of acceptable risks. This approach would allow users to develop resilient and adaptive strategies that recognise and reflect the specified levels of uncertainty. Strategies that are less brittle can have relatively less risk associated with their implementation.

Find out more

• See FAQ: Why have the 10th, 50th and 90th probability levels been chosen for the key findings?
• See User guidance: Projections of climate change (land)
• See User guidance: Projections of climate change (marine)
• See worked example for good practice: UKCP09 Guidance to perform an ecological assessment

Although the UKCP09 maps have the same sort of appearance as a weather map within the newspaper and maps available within UKCIP02, the nature of the image provided is quite different. For UKCIP02, the maps show the distribution of projected changes over the UK for a particular emission scenario, variable, temporal average, and 30-year time period. These UKCIP02 maps are similar to weather maps in terms of the image presented – a snapshot of the distribution of the climate at a particular time with all grid square values spatially coherent (i.e. a spatially coherent image of change).

UKCP09 maps are not presenting the same type of snapshot and users should take care that there interpretation of the UKCP09 maps is consistent with the information being presented by the image.

In the first place, maps available from UKCP09 show for each  grid square the value (e.g. projected temperature change) for which, based on the evidence considered, there is a given relative likelihood ( (e.g. 90% probability level) of the change being at or below that value. For example, consider a UKCP09 map showing grid square values at the 90% probability level of a change in summer mean temperature for a given 30-year time period and under a particular emission scenario. In this example, the value indicated in each grid square is that at which the evidence is suggesting there is a 90% probability of the projected mean summer temperature being below for that grid square, 30-year time period and emission scenario [see Section 4.3.1 Projections report].  Where there are higher values, the evidence considered is indicating that relatively higher values needed to be included in order to capture the same level of uncertainty.

Secondly, the values indicated in each grid square are a reflection of the contributions of the various sources of uncertainty for that grid square. These relative contributions will be different from one grid square to another [see Annex 2 of the projections report].  The model variants that provide the value at a given probability level may therefore be different for different grid squares and hence the values depicted on a given map are not necessarily spatially coherent.  As such, the UKCP09 maps are showing the changes in the depicted variable that needed to be considered for each grid squares in order to reflect the given probability level.

Finally, if the projected change of a variable at a particular location actually turned out to be the value at the 90% probability level, then this does not mean that the projected change for other variables at that location will also be that projected at the 90% probability level.  Nor does it mean that the change that occurs at other locations for that variable will also be at the value associated with the 90% probability level.  This aspect of relationships between projected changes becomes obvious when the joint probability between two variables (at the same or different locations) are considered.

Find out more  

• See User guidance: Preprepared maps and graphs

UKCP09 provides projections based on a 30-year averaging period.    As a result, climate for a specific month, season or year within a particular 30-year time period (e.g., 2020s) should not be compared with UKCP09 projections for that same time period.  One month, season or year does not make or break a 30-year averaging period.

An alternative way of looking at this is that one should not attempt to verify the projected outcomes for a given month, season or year as the real climate will only evolve along one particular trajectory. There is no opportunity to observe multiple possible future climates in a given month, season or year with which to compare the UKCP09 projections. See section 3.3 of the Projections report.

Find out more

• See FAQ: Is the 30 year baseline climate in the weather generator stationary or transient? Are the future projections stationary or transient?
• See FAQ: What are viable return periods to investigate?
• See User guidance: Projections of climate change (land)
• See User guidance: Projections of climate change (marine)
• Projections report: Box 1.3
• Projections report: Section 3.3 

The provided UKCP09 data sets (sampled data and CDF data) are not spatially consistent across adjacent grid squares and users should not average these to produce their own aggregated multi grid square areas. The climate represented by the projections has been derived based on climate models that produce outcomes that are spatially consistent across the grid squares. Further analysis of these model outcomes using the statistical framework to establish the UKCP09 data sets, however, results in data for each grid square that no longer reflect that spatial consistency.

Pre-defined aggregated areas of administration regions and river basin districts [see section 1.3.1 Projections report] and marine regions [see section 4.3.4 in the Projections report] have been specifically derived during the development of the probabilistic projections. These are the only aggregated areas larger than a single grid square that are available and that should be used.

Alternative Approaches 

Find out more

• See FAQ: Why can't I aggregate two grid cells together with the probabilistic data?
• See FAQ: How were the aggregated areas decided on and implemented?
• See User guidance: Projections of climate change (land)
• See User guidance: Projections of climate change (marine) 

The CDF data for a given grid square are not temporally consistent and should not be considered temporally linked from one temporal averaging period to the next (e.g. month to month, season to season, or one 30-year period to the next. As in the case of spatial consistency, the outcomes from the climate model are temporally consistent, however, in developing the CDF data this attribute is lost. As such, the CDF data available for several temporal averaging periods cannot be averaged to form a single CDF data set for a combined user-defined temporal averaging period (multiple months other than seasons provided or multiple seasons).

Alternative Approaches

Find out more

• See User guidance: CDF data

UKCP09 provides projections of climate change and climate for 30-year time periods. The projected climate outcomes are therefore an average for that 30-year period (monthly, seasonal and annual averages) and thus depict a stationary 30-year climate – there are no trends or changes beyond variability within a single time period. As such, users should not combine the data from multiple 30-year time periods to create a single data set to assess impact, risks and adaptation options across multiple 30-year time periods.

Alternative Approaches 

Find out more

• See FAQ: Is the 30 year baseline climate in the weather generator stationary or transient? Are the future projections stationary or transient?
• See User guidance: Projections of climate change (land)
• See User guidance: Projections of climate change (marine) 

The GIS Shapefiles were created using the CDF data.  As is outlined in the guidance, this data is a single variable data set and therefore does not include dependencies between variables.  Hence two variables overlayed in GIS will not relate to each other.

If users use the GIS shapefiles for representing one variable they need to be careful in interpretation of the resulting map.  These should not be interpreted in the same way as UKCIP02 maps.  For more information about this see the inappropriate use on "interpreting UKCP09 maps as weather maps" (above)

Alternative approaches

Find out more

• See User guidance: CDF data
• See User guidance: Sampled data