What's happening at IRRI
Thursday, February 04, 2016, 10:00am - 11:00am
LAI estimation using the Soil-Leaf-Canopy (SLC) radiative transfer model for irrigated rice: an exercise for Suphan Buri rice crop in the wet season of 2013
Dr. Nguyen Ha
The leaf area index (LAI) is an important biophysical parameter in terrestrial ecosystem models. LAI can be retrieved from remotely sensed data using either statistical or physical approach. While the former is mostly based on an empirical relationship between in-situ LAI and Vegetation Indices (i.e. NDVI), the later relies on the inversion of a canopy radiative transfer model (RTM). The inversion of canopy RTM can be sub-categorised into numerical optimisation look-up-table (LUT), and artificial neural network approaches.
In this study, an approach using LUT was used to reconstruct time series of LAI for rice growing areas in Suphanburi, Thailand. A specific RTM model known as Soil-Leaf-Canopy (SLC) was employed to generate LAI values for the rice fields in the wet season of 2013 using MODIS 8-day surface reflectance data (MOD09A1). Since the in-situ LAIs were measured at nearly every 25 days which almost didn’t coincide with MOD09A1 observations, interpolation had to be applied on SLC estimated 8-day LAIs in order to compare them with the ground measurement and to generate the daily LAI product. With Fourier time-series interpolation algorithm, estimated LAIs were found moderately correlated with in-situ LAIs (r= 0.74), though they were slightly underestimated. Meanwhile effort is still being made to use a more mechanistic function called Baret’s algorithm, which interpolates LAIs as an exponential function of cumulative daily temperature. The later function has the advantage of capturing the LAI evolutions throughout the whole season. The ultimate goal of this exercise is to be able to generate spatial LAI information that can be linked into the ORYZA crop growth model in order to be able to reconstruct historical yield at certain spatial aggregation level (such as at district level). Such information can be used to evaluate rice yield volatility for a given rice growing area and useful for the insurance component of the RIICE project.
SSD Conference Room, Drilon Hall