Despite the wide use of the Normalized Difference Vegetation Index (NDVI) in precision viticulture, there are no studies aimed at discriminating the contribution of plant biomass from that of leaf chlorophyll on canopy NDVI. Leaf area index (LAI) and leaf chlorophyll (Chl) were concomitantly monitored by ground measurements and projected canopy area (PCA), canopy volume (CV) and NDVI by high resolution UAV multispectral images in fully productive "Sangiovese" grapevines either grown in containers or in the field and subjected to different irrigation regimes over two consecutive years. NDVI values calculated from only vine canopy pixels (NDVIUAV) and NDVI obtained from mixed ground-canopy pixels (simulated Satellite NDVI, NDVISAT) were both evaluated as potential predictor of LAI and leaf Chl concentration.The seasonal patterns of LAI and leaf Chl concentration were affected by irrigation, showing differences depending on field vs container-grown conditions. In situations where a decoupling between LAI and leaf Chl occurred, NDVIUAV and NDVISAT showed different responses: NDVIUAV patterns strictly followed the leaf Chl ones, whereas NDVISAT was more affected by LAI. The coefficient of determination between NDVIUAV and leaf Chl ranged between 0.51 and 0.78, that between NDVIUAV and leaf Chl from 0.01 to 0.76, depending on the irrigation-growing conditions combination. NDVISAT was a better predictor of LAI (R-2=0.69) than NDVIUAV (R-2=0.42). In field-grown vines the relationships between NDVI (both UAV and SAT) and LAI was stronger than in potted ones. The relationships between LAI and PCA (R-2=0.44) or LAI and canopy volume (R-2=0.77) were both significant. The results allowed to confirm the two main hypotheses behind this experiment: i) leaf Chl concentration had a greater impact than LAI on NDVI values obtained from vine canopy pixels (NDVIUAV), whereas NDVISAT was more affected by LAI; ii) The canopy volume from UAV images was a better predictor of LAI than NDVI and the resulting relationship showed a better temporal stability.
The role of LAI and leaf chlorophyll on NDVI estimated by UAV in grapevine canopies
Caruso, G.;Palai, G.
;D(')Onofrio, C.;Gucci, R.
2023-01-01
Abstract
Despite the wide use of the Normalized Difference Vegetation Index (NDVI) in precision viticulture, there are no studies aimed at discriminating the contribution of plant biomass from that of leaf chlorophyll on canopy NDVI. Leaf area index (LAI) and leaf chlorophyll (Chl) were concomitantly monitored by ground measurements and projected canopy area (PCA), canopy volume (CV) and NDVI by high resolution UAV multispectral images in fully productive "Sangiovese" grapevines either grown in containers or in the field and subjected to different irrigation regimes over two consecutive years. NDVI values calculated from only vine canopy pixels (NDVIUAV) and NDVI obtained from mixed ground-canopy pixels (simulated Satellite NDVI, NDVISAT) were both evaluated as potential predictor of LAI and leaf Chl concentration.The seasonal patterns of LAI and leaf Chl concentration were affected by irrigation, showing differences depending on field vs container-grown conditions. In situations where a decoupling between LAI and leaf Chl occurred, NDVIUAV and NDVISAT showed different responses: NDVIUAV patterns strictly followed the leaf Chl ones, whereas NDVISAT was more affected by LAI. The coefficient of determination between NDVIUAV and leaf Chl ranged between 0.51 and 0.78, that between NDVIUAV and leaf Chl from 0.01 to 0.76, depending on the irrigation-growing conditions combination. NDVISAT was a better predictor of LAI (R-2=0.69) than NDVIUAV (R-2=0.42). In field-grown vines the relationships between NDVI (both UAV and SAT) and LAI was stronger than in potted ones. The relationships between LAI and PCA (R-2=0.44) or LAI and canopy volume (R-2=0.77) were both significant. The results allowed to confirm the two main hypotheses behind this experiment: i) leaf Chl concentration had a greater impact than LAI on NDVI values obtained from vine canopy pixels (NDVIUAV), whereas NDVISAT was more affected by LAI; ii) The canopy volume from UAV images was a better predictor of LAI than NDVI and the resulting relationship showed a better temporal stability.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
