On the Use of NDVI to Estimate LAI in Field Crops: Implementing a Conversion Equation Library

Abstract

The leaf area index (LAI) is a direct indicator of vegetation activity, and its relationship with the normalized difference vegetation index (NDVI) has been investigated in many research studies. Remote sensing makes available NDVI data over large areas, and researchers developed specific equations to derive the LAI from the NDVI, using empirical relationships grounded in field data collection. We conducted a literature search using “NDVI” AND “LAI” AND “crop” as the search string, focusing on the period 2017–2021. We reviewed the available equations to convert the NDVI into the LAI, aiming at (i) exploring the fields of application of an NDVI-based LAI, (ii) characterizing the mathematical relationships between the NDVI and LAI in the available equations, (iii) creating a software library with the retrieved methods, and (iv) releasing a publicly available software as a service, implementing these equations to foster their reuse by third parties. The literature search yielded 92 articles since 2017, where 139 equations were proposed. We analyzed the mathematical form of both the single equations and ensembles of the NDVI to LAI conversion methods, specific for crop, sensor, and biome. The characterization of the functions highlighted two main constraints when developing an NDVI-LAI conversion function: environmental conditions (i.e., water and light resource, land cover, and climate) and the availability of recurring data during the growing season. We found that the trend of an NDVI-LAI function is usually driven by the ecosystem water availability for the crop rather than by the crop type itself, as well as by the data availability; the data should be adequate in terms of the sample size and temporal resolution for reliably representing the phenomenon under investigation. Our study demonstrated that the choice of the NDVI-LAI equation (or ensemble of equations) should be driven by the trade-off between the scale of the investigation and data availability. The implementation of an extensible and reusable software library publicly queryable via API represents a valid mean to assist researchers in choosing the most suitable equations to perform an NDVI-LAI conversion.

Modelling Vegetation Health and Its Relation to Climate Conditions Using Copernicus Data in the City of Constance

“…An empirical approach was used through linear regression to establish the relationships between the vegetation indices and biophysical variables, specifically for LAI and FAPAR. Vegetation indices are a common indicator of vegetation status or growth model assimilation used to estimate LAI and FAPAR using remote sensing [24]. Five vegetation indices derived from Sentinel-2 in July 2018 were resampled to a coarser resolution of 300 m. The resampled images served as explanatory variables in the linear regression model of LAI and FAPAR from CDS.…”

Beyond Vegetation: A Review Unveiling Additional Insights into Agriculture and Forestry through the Application of Vegetation Indices

“…Strong linear relationship between the satellite-derived NDVI time series and the leaf area of the crop [57] Estimating corn LAI using hyperspectral reflectance data [58] Determine the radiation intercepted by the plant to estimate the LAI [59] Regional-scale method for accurately estimating rice LAI during the growing period [60] LAI estimation in semi-arid grasslands [61] Study of the trade-off between the scale of the research and the availability of data [62] Vegetation indices other than the NDVI to improve LAI estimations [63] However, when examining the scope of research, entries related to the NDVI and vegetation still vastly outnumber those associated with these recent concepts. Despite this disparity, a clear upward trend is observed in some areas, such as the articles related to the NDVI and climate change.…”

“…[61] focused their research in the semi-arid grassland of Inner Mongolia, formulating specific equations to estimate the leaf area index (LAI) within the typical vegetation range observed during the growing season. However, these equations vary from one area to another, and it is very important to select the appropriate NDVI-LAI equation, or a combination of equations, based on the trade-off between the scale of the research and the availability of data [62]. Nevertheless, the NDVI suffers from saturation at high density canopies and recently other VIs have been proposed to improve LAI estimations [63].…”

Updating of the Archival Large-Scale Soil Map Based on the Multitemporal Spectral Characteristics of the Bare Soil Surface Landsat Scenes

“…On the other hand, for the analysis of RSD, VI is most often used. Even on the topic (Special Issue) with an explicit indication of the BSS "Remote Sensing for Cropping Systems and Bare Soils Monitoring and Optimization", in most works a VIs were used [25][26][27][28]85,86]. There is a change of crops: corn [26], durum wheat [27], rice [28], and wheat [25], but the methods of processing RSD change little.…”

“…When processing RSD for crop production and soil mapping, the RED and NIR channels/bands are most often used [25][26][27][28]85,86]. The use of the RED and NIR channels goes back to the description of the "tasseled cap" [87,88].…”