Literature Review
1.3. NDVI-LAI Relationship
Several studies have shown that Normalized Difference Vegetation Index (NDVI) and Leaf Area Index (LAI) have been used for yield estimation and prediction. Both these parameters give the vegetation status of the area. It has been observed that NDVI and LAI provide the relation with yield that predicts the yield fluctuations. Many models have been tried in this regards. The LAI is an important vegetation canopy characterization. It is one parameter for crop growth models. The major interest of remote sensing LAI retrievals took place since the launch of earth observation satellite. It is indeed difficult to build accurate and robust retrieval models due to the complexity of relationship between retrieved signal and LAI. But MODIS LAI could be developed and validated (Guissard, 2005). Since NDVI and LAI are obtained from satellite data (Gardner et.al., 1991), the distinction of vegetation and quantitative assessment for growth has become possible. Since plant canopy is composed of leaves, which is a direct source of the energy-matter interactions that are observed by earth-observing remote sensing systems, LAI has been an attractive variable of interest in vegetative remote sensing. However, large portion of such studies to estimate LAI using NDVI were dealing with semi-arid vegetation and agricultural systems where the canopy closure is less than 100%. Studies have also been carried out where the linear models were evaluated using NDVI and field LAI. Cross validation procedure was used to assess the prediction power of the regression models. The study demonstrates that for hyperspectral image data, linear regression models can be applied to quantify LAI with good accuracy (Clement et.al, 2004). LAI – NDVI relationship was also used for estimating the evapotranspiration where the relationships between the NDVI – LAI were adjusted by potential model. 57 – 72% variance of NDVI was explained by the LAI (Xavier et.al, 2004). Research has also been done where the correlation has been developed between remote sensing data and strawberry growth and yield. Regression analysis was then used to determine the relationship between canopy size and the NDVI from the aerial image. The results showed positive correlation between the two sets of data (r2=0.918). Studies have also been carried out to find the correlation between the maximum latewood density of annual tree rings and NDVI based estimates of forest productivity (Darrigo et.al., 2000). This study showed positive correlations between maximum latewood density data and NDVI based net primary productivity (NPP). Remote sensing from satellite and aerial platforms has been used to detect nutrient status of plants (Blackmer et al., 1996; Wright et al., 2000) and estimate yield (Shanahan et al., 2001; Labus et al., 2002). Some experiments have already been carried out on wheat plants. Most of the yield prediction estimates on larger scales use a simple linear regression model derived from the normalized difference vegetation index (NDVI) with imagery collected during the wheat grain filling period (Benedetti and Rossini, 1993; Doraiswamy and Cook, 1995). Correlation coefficients derived from remote sensing range between 0.7 to 0.83 (r2) in uniform crop patterns (Raun, et al., 2000) with the expectation of better correlation as spatial and spectral resolution increases and algorithms improve. From the above discussion it is quite clear that correlation and regression between NDVI and LAI could help in linear model development which could further help in estimating and predicting yield.