Climate Exchange

Data rescue at Météo-France

Advocating for data rescue and convincing a wide range of actors is

a necessary step to ensure that all stakeholders bring resources and

work together, beginning with the recognition that old documents

are a major community asset and a rich source of knowledge for

coping with climate change effects. Data rescue at Météo-France is

no longer a research activity. Météo-France’s climatologists put a

special emphasis on archives, aiming to preserve old documents still

kept in-house in good condition. It is now necessary to look again

at these documents, as climatologists who had typically worked at

monthly timescales a few years ago must now reprocess daily or

sub-daily data as they focus on extreme events. There is also a need

to search in various archiving agencies for unknown documents and

retained information.

Climatologists first of all want to retrieve observations from old

documents. Enriching the data banks is not the sole goal. They

definitely dream of raw data but also of the compulsory metadata,

explaining the measurements’ conditions and thus enabling a

[

] 269

apacity

evelopment

higher quality of existing series and derived diagnos-

tics. A primary goal for Météo-France is to improve

the quality and length, but also the spatial and param-

eter coverage of the existing series, used for detection

and attribution of climate change. Data rescue feeds

analysis and reanalysis. Preservation and recovery are

the first steps, followed by a stringent homogenization

procedure. Homogenization is a geostatistical evalua-

tion of the series: it detects shifts due to the evolution

of measurement conditions. While the tradition was a

day-to-day quality control of the observation, clima-

tologists now check the quality of the entire series.

They shed light on the vital need for metadata related

to the measures. For example, which instruments were

used? When was such a new model put in place? Did

the surrounding conditions evolve?

Various other aims motivate data rescue. One should

first of all mention political motivation: society must be

sure that climatologists have made all possible efforts

to improve the quality of the delivered diagnosis, and

exploited all potentially available information gathered

by the meteorological services. Facts can be criticized

and they need to be as strong as possible. Secondly, and

not least, our duty is to provide climate research with

all elements that would help scientists to better under-

stand the Earth’s climate system: long series, facts and

figures, especially on extremes or gaps, such as wars,

ex-colonies or upper-air, all being targeted as primary

goals for global reanalysis projects.

The originality of the current effort at Météo-France

lies in the transfer of such activities from a research

mode to operations. Meteorologists in every local

service, including in overseas territories, are commit-

ted to these tasks. Homogenization, which requires

skills and was originally a research effort, is becom-

ing a more widely spread skill for climatologists. This

opens their horizons to climate science, and they

gradually move from data control to climate analysis,

trends and extremes, enabling them to fruitfully inter-

act with climate scientists and enriching their dialogue

for mutual benefit. Data rescue and homogenization is

therefore also a strategy for improving skills and widen-

ing horizons. This is true in the meteorological service

and it appears to be true outside. Climate services are

not so far away when it comes to educating partners in

impact communities.

Cross-cutting partnerships

Data rescue actions are carried on in our offices,

where a lot of archives are kept. The organization of

the preservation and archive automatically opened

the meteorological community to partners, mainly

archivists. This partnership is critical. Not only does

it bring knowledge on archiving, but it also opens

connections with partners from communities like

historians or heritage specialists who enrich climate

knowledge. Documents are also searched for in various

archive agencies and libraries, having national, local or

thematic duties. Documents are located, identified and,

Source: Météo-France and Archives Nationales

A chart of temperatures for 07:00 on 17 June 1924 – The values on such

charts can help today’s climatologists to understand rapidly changing

climate events