Infering meteorological information at different scales from several sources of data
Résumé
One of the major challenges to monitor, foresee and anticipate the landscape evolution is to
understand the climate change geographical effects (Aspinal, 2012) on different, especially local,
scales (Martin et al., 2013; Barry & Blanken, 2016). For instance, measuring the evolution of
temperature is possible, one the one hand, using a series of (often spatially accurate, but irregular)
sensors spread over a given territory (e.g. a watershed). On another hand, we have today a large
access to climate models, providing meteorological projections at a certain (generally coarse, but
regular) spatial and temporal granularity. From both those types of data, scientist can infer
information at different (nested) scales, applying upscaling (by spatial data aggregation) or
downscaling (assuming some disaggregation hypothesis) processes.
This work deals with this issue: how can we infer a reliable spatial information from meteorological
data provided from two different levels and methodologies? More precisely, can we provide relevant
estimations on climate drivers at different scales, in the particular case of meteorological data
(illustrated by temperature)? To answer to these two questions, we face two problems: (i) how to
combine these two sources of data (ii) and how to deal with the Ecological Effect (Holt, 1996; King,
1997) or the Modifiable Areal Unit Problem (Openshaw, 1984), that may strongly impair the
estimate reliability and usability to forecast the likely climate landscapes for the future?
In this work, we present recent results obtained by studying temperatures measurements from a set of
meteorological stations and the ALADIN model grid on long time series in the French southern
region of Provence Alpes Côte d’Azur. Using those two sources of data, we aggregate
temperature values and observe their variation through different administrative territorial partitions
(somehow French delineations under the regional scale NUTS2 : “départements”, “arrondissements”,
“cantons”, “communes” and “EPCI”, i.e. groups of communes).
This leads us to draw what we call “scalograms” which plot average or median
temperatures according to the different nested levels of scale. Those are provided for both the
gridded ALADIN model and the series of local meteorological stations, and compared. We notice
some differences in the estimations that show the necessary caution to pay for generating
meteorological data using a multiple scale approach. A method, published a few year ago, based on
spatial random permutation (Josselin et al., 2008) and generalized to any data (Josselin et al., 2023),
is applied on this kind of climate data to mitigate the change of support problem and to improve the
data reliability to potentially characterize more accurately the landscapes at different scales.
References
Aspinal R. (Ed.). (2012). Geography of climate change. Routeledge, Taylon Fancis.
Barry R. G., Blanken P. D. (Eds.). (2016). Microclimate and local climate. Cambridge University Press.
Holt D., Steel D., Tranmer M., Wrigley N. (1996). Aggregation and ecological effects in geographically based
data. Geographical Analysis, vol. 28, p. 244-261.
Josselin D., Mahfoud I., Fady B. (2008). Impact of a change of support on the assessment of biodiversity with
shannon entropy. In Spatial Data Handling, SDH’2008", pp. 109-131. Montpellier, June, 23-25.
Josselin D., Blanke D., Coulon M., Boulay G., Casanova Enault L., Peris A., Le Brun P., Lecourt T. (to be
published, 2023). Incertitudes liées aux échelles d'estimation des prix immobiliers. In L’imperfection des
données géographiques. Tome 2. (Eds: M. Batton-Hubert, E. Desjardin, F. Pinet), ISTE-Wiley
King G. (1997). A solution to the ecological inference problem. Reconstructing individual behaviour from
aggregate data. Princeton University Press.
Martin N., Carrega P., Adnès C. (2013). Downscaling à fine résolution spatiale des températures actuelles et
futures par modélisation statistique des sorties ALADIN-climat sur les Alpes-Maritimes (France).
Climatologie, Association internationale de climatologie, pp.51-72.
Openshaw S. (1984). The modifiable areal unit problem. Norwich: Geo Books, CATMOG 38.