Investigating the impact of leaf area index temporal variability on soil moisture predictions using remote sensing vegetation data

- Constant LAI is sufficient to simulate daily soil moisture.
- Land surface model's sensitivity to average LAI determines the influence of LAI seasonality.
- Impact of LAI seasonality on reproducing soil moisture is soil-dependent.
- Calibrating a constant LAI can reduce the impact of LAI seasonality.

SummaryThe impact of leaf area index (LAI) seasonality on three-year (2005-2007) daily soil moisture predictions was investigated for two different land surface models (IBIS and HYDRUS) at the Stanley semi-arid grassland field site. Three daily LAI time series derived from different empirical NDVI-LAI relationships using the MODIS NDVI data were used in the analysis. Calibration results from both models consistently suggested that an average LAI over time, rather than a time varying daily LAI, was sufficient to reproduce daily soil moisture at our site. We did, however, find that the sensitivity of the impact of LAI time variability on soil moisture estimation was a function of soil parameters. The influence of LAI time variation on the soil moisture simulations is controlled by the sensitivity of modelled soil moisture to the average LAI values over that period, and soil parameters affected the sensitivity of the model to LAI. Those parameter sets that were most sensitive to the long-term mean LAI were also those that were the most sensitive to the time variability. In our case, model calibrations using a constant LAI adjusted the soil parameters to reduce the impact of LAI variability. Results also suggested that the LAI variability could be significant if the varying LAI approached a very low level (i.e. LAI < 1) for a significant proportion of the simulation period. This is most likely to be the case for short grasses in grasslands.