Methodological considerations on local temperature zoning: how many sensors and what interpolation method could be used? A case study in Bourgogne (France)

Abstract

Due to the development of connected and moderately cheap temperature sensors, climate zoning within vineyards is now more accessible to both the research community and the wine industry. However, the methodological framework for implementing climate zoning remains unclear. In order to clarify the methodology for climate mapping at the local level, we explore here the impact of sensor network density on the accuracy of climate mapping at the local level. Within an 800-hectare wine-growing area in Burgundy (France), 112 temperature sensors were deployed 60 cm above the ground in vineyard plots from March 2023 to September 2024. Spatial interpolation of daily minimum (Tmin) and maximum (Tmax) temperatures was compared for seven methods, first with all sensors deployed, then by reducing the number of sensors through stratified sampling based on terrain characteristics. Our results show that below 40 sensors (5 sensors/km²), spatial interpolation has more than a modeling day out of two an efficiency of less than 0.4, indicating poor performance. Between 5 and 10 sensors/km², random forest interpolation, calibrated by machine learning, performs best, especially for maximum temperatures, where interpolation is most inaccurate. With 112 sensors (14/km²), the regression-kriging method achieves the highest accuracy. The average RMSE reaches 0.4°C for Tmin and 0.56°C for Tmax. These results, as well as the impact of interpolation errors on various agroclimatic indices (average temperature during the growing season, heat stress indices), provide useful information for the implementation of climate zoning of wine-growing terroirs.

References

Bois, B. (2024). Climate diversity of vitivinicultural terroirs at local scale level. In R. Lopez Plantey (Ed.), Proceedings of the 15th International Terroir Congress (pp. 88–65). UNCUYO. https://congressterroir.com.ar/15%20International%20Terroir%20Congress_PROCEEDINGS.pdf

Bonnefoy, C., Quénol, H., Barbeau, G., & Madelin, M. (2010). Analyse climatique à l’échelle des Coteaux du Layon. Proceedings of the 8th International Terroir Congress, 3109–3114. https://ives-openscience.eu/8745/

de Rességuier, L., Mary, S., Le Roux, R., Petitjean, T., Quénol, H., & van Leeuwen, C. (2020). Temperature Variability at Local Scale in the Bordeaux Area. Relations With Environmental Factors and Impact on Vine Phenology. Frontiers in Plant Science, 11. https://doi.org/10.3389/fpls.2020.00515

Gruijter, J. de, Brus, D. J., Bierkens, M. F. P., & Knotters, M. (2010). Sampling for Natural Resource Monitoring (1st ed. Softcover of orig. ed. 2006). Springer-Verlag Berlin and Heidelberg GmbH & Co. K.

Hengl, T. (2006). Finding the right pixel size. Computers & Geosciences, 32(9), 1283–1298. https://doi.org/10.1016/j.cageo.2005.11.008

Jacquet, A., & Morlat, R. (1997). Caractérisation de la variabilité climatique des terroirs viticoles en val de Loire. Influence du paysage et des facteurs physiques du milieu. Agronomie, 17, 465–480.

Le Roux, R., de Rességuier, L., Corpetti, T., Jégou, N., Madelin, M., van Leeuwen, C., & Quénol, H. (2017). Comparison of two fine scale spatial models for mapping temperatures inside winegrowing areas. Agricultural and Forest Meteorology, 247, 159–169. https://doi.org/10.1016/j.agrformet.2017.07.020

Madelin, M., Bois, B., & Quénol, H. (2014). Variabilité topoclimatique et phénologique des terroirs de la montagne de Corton (Bourgogne). In H. Quénol (Ed.), Changement climatique et terroirs viticoles (pp. 215–228). Lavoisier.

Nash, J. E., & Sutcliffe, J. V. (1970). River flow forecasting through conceptual models part I — A discussion of principles. Journal of Hydrology, 10(3), 282–290. https://doi.org/10.1016/0022-1694(70)90255-6

Perreault, S., El Alem, A., Chokmani, K., & Cambouris, A. N. (2022). Development of Pedotransfer Functions to Predict Soil Physical Properties in Southern Quebec (Canada). Agronomy, 12(2), 526. https://doi.org/10.3390/agronomy12020526