Submitted to: Field Crops Research
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/22/2007
Publication Date: 5/31/2007
Citation: Wu, J., Wang, D., Bauer, M. 2007. Assessing Broadband Vegetation Indices And QuickBird Data In Estimating Leaf Area Index Of Corn And Potato Canopies. Field Crops Research. Interpretive Summary: Remotely sensed measurements provide great potential for monitoring crop activities, although remote sensing cannot directly measure agronomic variables such as leaf area index (LAI) and crop biomass. An improved method was developed to evaluate the efficiency of commonly used broadband VIs in estimating LAI. The method took into account all three factors that affect the efficiency of VIs, i.e., stabilities, sensitivities, and dynamic ranges. Based on field measurements made in three growing seasons, we analyzed LAI noise inherent in each VI–LAI function and examined the capacity of QuickBird data for monitoring absolute LAI values and spatial variabilities. QuickBird image-estimated LAI using the inverted MSAVI–LAI relationship agreed well with ground measurements in both absolute values and spatial variability. The evaluation and validation were based on corn and potato, and the parameters of MSAVI–LAI were relatively consistent for different crops. Thus, the results should have implications for other crops. For the purpose of agricultural applications at the field scale, the semi-empirical MSAVI–LAI relationships are reasonably efficient for estimating LAI with satisfactory absolute values and spatial variability.
Technical Abstract: Leaf area index (LAI) is a key biophysical variable that can be used to derive agronomic information for field management and yield prediction. In the context of applying broadband and high spatial resolution satellite sensor data to agricultural applications at the field scale, an improved method was developed to evaluate commonly used broadband vegetation indices (VIs) for the estimation of LAI with VI–LAI relationships. The evaluation was based on direct measurement of corn and potato canopies and on QuickBird multispectral images acquired in three growing seasons. The selected VIs were correlated strongly with LAI but with different efficiencies for LAI estimation as a result of the differences in the stabilities, the sensitivities, and the dynamic ranges. Analysis of error propagation showed that LAI noise inherent in each VI–LAI function generally increased with increasing LAI and the efficiency of most VIs was low at high LAI levels. Among selected VIs, the modified soil-adjusted vegetation index (MSAVI) was the best LAI estimator with the largest dynamic range and the highest sensitivity and overall efficiency for both crops. QuickBird image-estimated LAI with MSAVI–LAI relationships agreed well with ground-measured LAI with the root-mean-square-error of 0.63 and 0.79 for corn and potato canopies, respectively. LAI estimated from the high spatial resolution pixel data exhibited spatial variability similar to the ground plot measurements. For field scale agricultural applications, MSAVI–LAI relationships are easy-to-apply and reasonably accurate for estimating LAI.