Assessment of vegetation productivity in the Umfolozi catchment using Leaf Area Index (LAI) derived from SPOT 6 image.
Abstract
Around the world, rural areas rely on the natural resources for their sustenance. These
include grazing lands for livestock production and fuel wood harvesting for heating
and cooking, as well as for medicinal purposes. These natural resources are barely
managed in rural areas which exacerbate the challenge of land degradation due to unsustainable
overgrazing and fuel wood collection. Land degradation has been identified
as one of the key global problems are the root cause of poverty, food insecurity
and malnutrition. In South Africa, the uMfolozi catchment is very vulnerable to disturbance
due to slow ecological recovery, growing human populations and episodic
droughts. Leaf Area Index (LAI), defined as one half the total green leaves per unit
ground surface area, is an inventory of the plant green leaves that defines the actual
size of the interface between the vegetation and the atmosphere. Thus, LAI spatial data
could serve as an indicator of vegetation productivity. The main aim of the study is
to estimate LAI as an indicator of vegetation productivity using remotely sensed data.
First, field collected LAI were used to assess LAI models derived from various vegetation
indices and bands. Secondly, multivariate statistics were used to combine bands
and indices in estimating LAI. Combining reflectance at various bands and vegetation
indices yielded higher estimation accuracy of LAI (Bootstrapped: R² = 0.71, RMSE =
0.92) as compared to using individual bands or indices. Furthermore the study found
that environmental variables such as slope, Digital Elevation Model (DEM) and annual
mean temperature significantly influenced the spatial distribution of LAI. There is a
scope to estimate LAI empirically using bands and vegetation indices which are more
site and data specific, but the study further recommends the use of physically-based
models which are known to be robust. In conclusion, estimation of LAI is possible using
remote sensing derived variables combined with multivariate statistical techniques,
which is critical for assessing vegetation productivity.