Using Normalized Difference Vegetation Index (NDVI) as a Proxy for Plant Size in Water Use Models to Facilitate Precision Irrigation

Precise irrigation of greenhouse crops based on plant water needs not only allows for optimal plant growth, but also conserves water and alleviates environmental pollution from fertilizer and pesticide runoff. A thorough understanding of crop-specific water requirements is essential for more efficient irrigation. However, plant water use changes on a daily basis, driven by variations in environmental conditions as well as increases in plant size over time. While environmental conditions are relatively easy to measure, direct determination of plant size is often destructive and time consuming. Remote sensing of vegetation indices, such as the normalized difference vegetation index (NDVI), provides a continuous and non-destructive method to estimate canopy size for use in water use models. The objective of this study was to develop quantitative models that predict daily water use (DWU) of four bedding plant species based on environmental factors and NDVI, a proxy for plant size. In addition, we also determined the feasibility of using NDVI in place of ‘crop coefficients’ (the ratio of crop DWU to reference evapotranspiration) that are commonly used in agronomic applications. The NDVI increased linearly over time during vegetative growth, until canopy closure was nearly complete. However, flowers reflect light at the two wavebands of interest (centered at 650 nm & 810 nm, respectively) differently from green leaves.  The NDVI measured from three of the plant species that flowered within the course of the study showed little increase or declined gradually when flowers started to form and shade the leaves. The crop coefficients increased linearly over time as plants grew, but NDVI increased quadratically with increasing crop coefficient, indicating decreasing sensitivity of NDVI to increasing plant size.  Nevertheless, multiple linear regression models developed using reference evapotranspiration and NDVI were able to explained 80% to 96% of variation in DWU of each individual species, suggesting that NDVI may be used as a reliable proxy for plant size.