Water status of trees in real urban environment: measurements and understanding at the individual tree scale, modeling based on very high-resolution satellite imagery. The case of Dijon

Started in november 2022

Funding: French National Research Institute for Sustainable Development (IRD), Paris

Supervisors: Nadège Martiny (Dijon) & Christian Hartmann (Paris)

Defense the 18 december 2025

Abstract

Urban trees play a major role in mitigating the effects of climate change, particularly in relation to droughts and urban heat islands. However, their water functioning remains poorly characterized under real urban conditions. This thesis aims to characterize the water status of urban trees, identify the factors that control it, and assess the potential of very high spatial resolution (VHSR) remote sensing for its modeling at both individual and seasonal scales. The study was conducted in Dijon (France) on three tree species widely planted in Western Europe (Tilia sp., Acer sp., and Aesculus sp.) and relied on three complementary approaches: (1) high-precision dendrometric monitoring of eleven instrumented trees over more than one year, allowing the calculation of two reactive indicators of water status, the Tree Water Deficit (TWD) and the Maximum Daily Shrinkage (MDS); (2) a terrestrial LiDAR survey of 600 trees to estimate Leaf Area Index (LAI), used here as an integrative proxy of water status at the end of the growing season; and (3) the extraction of vegetation indices (VIs) from VHSR satellite imagery to evaluate their ability to capture variations in these metrics. Results show that TWD is a robust indicator of water deficit, while MDS mainly reflects daily physiological activity. Atmospheric conditions are major determinants of water status, with vapor pressure deficit (VPD) in summer and relative humidity in spring strongly influencing tree water dynamics. Soil sealing exacerbates water constraints: trees planted in pits displayed consistently higher deficits than those in tree strips, themselves more stressed than park trees. Marked interspecific contrasts were also observed: Tilia × euchlora exhibited an anisohydric strategy with strong water potential fluctuations, while Acer platanoides adopted a more stable isohydric behavior. LAI proved sensitive to both soil sealing and tree health status across species, confirming its value as an integrative metric. Spectral index analysis highlighted the potential of VIs to reflect water status, particularly in summer. Two groups emerged: NDVI–NDRE–SIPI on the one hand, and EVI–MSAVI2 on the other, the latter incorporating an explicit soil correction. While explanatory models of TWD showed limited interannual transferability, significant correlations confirmed the relevance of VIs to capture water functioning in urban contexts. At the integrative scale, a machine learning model combining spectral indices with biological, health, and environmental variables explained 19.4% of the spatial variance of LAI, with NDVI performing slightly better than MSAVI2. In summary, the water status of urban trees is multifactorial, controlled by the interaction of atmospheric conditions, soil sealing, and biological traits. VHSR remote sensing shows strong potential to capture hydric indicators, provided it is coupled with multivariate approaches and the integration of complementary variables. These findings pave the way toward a functional urban remote sensing framework, able to inform tree management and support urban resilience strategies.

Keywords

urban trees, water status, imperviousness, very high resolution imagery, dendrometry, leaf area index

Jury

Thomas Corpetti, université de Rennes – reviewer
Stéphane Herbette, Université Clermont Auvergne – reviewer
Sophie Herpin, Institut Agro Rennes-Angers – examiner
Jean-François Léon, Université de Toulouse – examiner
Anne Puissant, Université de Strasbourg – examiner
Nadège Martiny, Université Bourgogne Europe – supervisor
Christian Hartmann, IRD – cosupervisor
Thomas Bur, Société Urbasense – invited
Nicolas Marilleau, IRD – invited

extrait:

lien_externe:

titre:

Végétation en zone urbaine : rôle sur les cycles de l’eau et du carbone, impact sur la qualité de l’air et le rafraîchissement à micro-échelles. Le cas de Dijon Métropole

date_de_debut_these:

novembre 2022

nom:

Canovas

date_de_debut_these_numerique:

202211

kc_data:

a:8:{i:0;s:0:"";s:4:"mode";s:0:"";s:3:"css";s:0:"";s:9:"max_width";s:0:"";s:7:"classes";s:0:"";s:9:"thumbnail";s:0:"";s:9:"collapsed";s:0:"";s:9:"optimized";s:0:"";}

kc_raw_content:

Water status of trees in real urban environment: measurements and understanding at the individual tree scale, modeling based on very high-resolution satellite imagery. The case of Dijon

Started in november 2022

Funding: French National Research Institute for Sustainable Development (IRD), Paris

Supervisors: Nadège Martiny (Dijon) & Christian Hartmann (Paris)

Defense the 18 december 2025

 

Abstract

Urban trees play a major role in mitigating the effects of climate change, particularly in relation to droughts and urban heat islands. However, their water functioning remains poorly characterized under real urban conditions. This thesis aims to characterize the water status of urban trees, identify the factors that control it, and assess the potential of very high spatial resolution (VHSR) remote sensing for its modeling at both individual and seasonal scales. The study was conducted in Dijon (France) on three tree species widely planted in Western Europe (Tilia sp., Acer sp., and Aesculus sp.) and relied on three complementary approaches: (1) high-precision dendrometric monitoring of eleven instrumented trees over more than one year, allowing the calculation of two reactive indicators of water status, the Tree Water Deficit (TWD) and the Maximum Daily Shrinkage (MDS); (2) a terrestrial LiDAR survey of 600 trees to estimate Leaf Area Index (LAI), used here as an integrative proxy of water status at the end of the growing season; and (3) the extraction of vegetation indices (VIs) from VHSR satellite imagery to evaluate their ability to capture variations in these metrics. Results show that TWD is a robust indicator of water deficit, while MDS mainly reflects daily physiological activity. Atmospheric conditions are major determinants of water status, with vapor pressure deficit (VPD) in summer and relative humidity in spring strongly influencing tree water dynamics. Soil sealing exacerbates water constraints: trees planted in pits displayed consistently higher deficits than those in tree strips, themselves more stressed than park trees. Marked interspecific contrasts were also observed: Tilia × euchlora exhibited an anisohydric strategy with strong water potential fluctuations, while Acer platanoides adopted a more stable isohydric behavior. LAI proved sensitive to both soil sealing and tree health status across species, confirming its value as an integrative metric. Spectral index analysis highlighted the potential of VIs to reflect water status, particularly in summer. Two groups emerged: NDVI–NDRE–SIPI on the one hand, and EVI–MSAVI2 on the other, the latter incorporating an explicit soil correction. While explanatory models of TWD showed limited interannual transferability, significant correlations confirmed the relevance of VIs to capture water functioning in urban contexts. At the integrative scale, a machine learning model combining spectral indices with biological, health, and environmental variables explained 19.4% of the spatial variance of LAI, with NDVI performing slightly better than MSAVI2. In summary, the water status of urban trees is multifactorial, controlled by the interaction of atmospheric conditions, soil sealing, and biological traits. VHSR remote sensing shows strong potential to capture hydric indicators, provided it is coupled with multivariate approaches and the integration of complementary variables. These findings pave the way toward a functional urban remote sensing framework, able to inform tree management and support urban resilience strategies.

 

Keywords

urban trees, water status, imperviousness, very high resolution imagery, dendrometry, leaf area index

 

Jury

Thomas Corpetti, université de Rennes - reviewer
Stéphane Herbette, Université Clermont Auvergne - reviewer
Sophie Herpin, Institut Agro Rennes-Angers - examiner
Jean-François Léon, Université de Toulouse - examiner
Anne Puissant, Université de Strasbourg - examiner
Nadège Martiny, Université Bourgogne Europe - supervisor
Christian Hartmann, IRD - cosupervisor
Thomas Bur, Société Urbasense - invited
Nicolas Marilleau, IRD - invited