Derivation of global empirical relationship between climate variables and leaf area index for the parametrization of large-scale hydrological models

Abstract

A novel model was developed to find the dynamic relationship between the climate variables and regional leaf area index (LAI). The NASA MODIS datasets were used for the LAI and Land Use Data. The Climate Research Unit (CRU) provided meteorological data. The model was developed in Python programming language with the NumPy and Gdal packages. The examination of the climatic variables reveals that temperature, precipitation, vapour pressure, and potential evapotranspiration influence LAI the most on a regional and global scale. The local analysis was performed for five locations at the Temperate and Tropical Zone of the Köppen’s Classification. The results revel relatively good prediction. By taking additional parameters in the model (for instance radiation) the model could improve its performance at the local scale. The global analysis of the most influential climate variable shows strong connection between potential evapotranspiration and precipitation at the equator. The temperature and vapour pressure increase in importance with northern and southern direction. The climate variables, that influence the regional LAI, depend more on the climate zone then plant functional type. The satellite data is very sensitive and LAI measurements are dependent on the correct interpretation of the data and on accurate filtering measurements errors (clouds). To improve the predictive equation for the LAI estimation, climate variables should have a memory of the previous months.