Assessing Climatic Stress in Vegetation: A Statistical-Driven Approach to Predict Thermal “Degradation” Parameters via Passive Thermography

Vertical green structures (VGS) increase green spaces in compact urban areas offering benefits as an improved thermal insulation of the building envelope, and a decrease of the Urban Heat Island effect. Efficient species selection and quantitative evaluation of species’ reaction to hygro-thermal variations are needed to optimize VGS benefits. This study proposes a novel methodology to discriminate the thermal behavior between healthy and unhealthy leaves under hygrothermal stress. To this aim, the surface temperature on region of interests (ROIs) of target leaves of the Heuchera villosa species is measured using passive thermography (pIRT). Besides, an additively manufactured leaf is used as a reference shape for assessing geometrical changes between healthy and unhealthy leaves. The thermal data analysis procedure extracts a vegetative health index based on the third-order centered momentum of thermal data distribution, encompassing temperature anomalies, and geometric variations. The study demonstrates high potential to be further extended to a data-driven approach to automatically and non-destructively detect the healthy and unhealthy status of vegetation in VGS by pIRT. The findings contribute to advancing understanding of vegetative responses to environmental stressors and provide insights for effective monitoring and management of VGS.