Climate Sense

[

] 258

dAptAtion

And

itigAtion

trAtegieS

rather than to get the best possible information from the

climatic point of view. Hydropower production forecast-

ing on an annual time-scale is, for example, a key element

of the French electricity management system. Current

practices, based on the use of historical data, could be

easily improved using climate forecasts and downscaling

methods, by simply discriminating between dry/normal/

wet periods a few weeks or months in advance. Moreover,

long range forecasts are often more accurate in tropical

areas, and could therefore be profitably used in develop-

ing countries to forecast water resource, in particular, as

is already the case in western Africa

and South America.

This could help such countries to better manage their

water resources and their electricity production.

A second major issue is the use of probabilistic infor-

mation, and its communication to end users. Ensemble

(probabilistic) forecasts are particularly well suited to

answer risk management needs. But end users do not

generally feel comfortable using probabilistic weather and

climate forecasts, whereas products and services already

exist that may answer current identified needs with little

adaptation. With this in mind, National Meteorological

and Hydrological Services (NMHSs) have, over the last few

years, developed alert systems (for example, in Europe,

Meteoalarm and Carte de Vigilance from Météo-France).

According to the prevention principle these services gener-

ally issue more false alarms than was previously the case,

which could be seen as a drawback. But early warnings of

severe weather risks allow a quicker reaction andmore effi-

cient organization of emergency units, for example. They

also enable better communication with national or local

authorities, reducing the exposure of people and goods

to hazards, and hence the consequences of the event, as

taking preventive measures is generally cheaper than treat-

ing problems, even if events finally do not occur. Over

the next few years special emphasis should therefore be

placed on the importance of communication. In particular,

emphasis should be placed on fostering communication

of uncertainties and discussions between NMHSs and

end users. Different kinds of decision can be taken and

different levels of reaction can be activated, depending

on the forecasted event itself, but also on the uncertainty

associated with the forecast. This could notably drive

adaptations in the decision processes schemes for differ-

ent sectors, cross-sectors and authorities.

Adaptation to climate change and mitigation strategies

is likely to prompt the development of renewable energy

sources, such as electricity production, over the coming

decades. This reveals two additional and important needs:

the first one concerns the availability of weather and

climate data and forecasts that are not currently produced

and proposed routinely by NMHSs (for example, wind

speed at wind turbines’ hub height – around 100 metres).

This problem could be easily overcome with an increased

dialogue between providers and users. The second need

concerns climate projections for the next 30 years, which

corresponds to the lifetime of renewable productionmeans,

in order to build the best business plans and guide the best

investment decisions. Current projects on decadal predic-

end of the century, towards which the cooling systems are to be

dimensioned. Where the climate change impacts on a given param-

eter are not yet reliably known, a minimal value – taking present

climate change impact knowledge into account – is assumed, and

actions are planned correspondingly in order to allow future adapta-

tion measures during the running life of the installation.

Temperature increase is now included in the long-term electricity

consumption forecasts. Moreover, research is conducted in order to

evaluate and maximize the use of medium to long-term forecasts.

Real and future energy needs using climate information services

The energy sector is probably one of the most important users of

weather, water and climate data and forecasts. However, there are still

many gaps to fill in order to better manage energy production and

distribution systems in developed countries, and to build efficient and

sustainable systems in developing countries.

One major issue is the

development and improvement of forecasting systems at lead times

from two weeks to one or several years – the so-called ‘seamless predic-

tion systems’.

Monthly, seasonal and decadal climate predictions go a

good way to answer some of the needs, but there are still many chal-

lenges to face. These forecasting systems need to be notably improved

in order to be able to deliver reliable information to the users. Even

if climate predictability is low over mid-latitudes, current prediction

systems show positive scores for some variables, seasons and types

of events. These performances may be superior to the current energy

sector’s practices and closer collaborations between users and providers

are the only way to identify possible improvements in the electricity

sector applications and decision tools. They may also guide model

development in a positive way. In particular, special attention should

be paid to both the adaptation of such forecasts to the users’ needs – for

example using downscaling methods which allow the transformation of

large scale models outputs to local users’ needs for application models

– and on the consistency of the information delivered at different time-

scales. The problem here is getting the suitable information adapted

to the decision processes, which differ according to the time horizon,

Villagers and solar panels in Towé & Houéddo badgi (Benin)

Image: EDF Mediatheque/Riffet Daniel