By Jacob Chuma Beatus
Abstract
The light use efficiency (LUE) approach is often used with remotely sensed data products and meteorological data to estimate net primary production (NPP) from local to global scales. To estimate local NPP, detailed information of spatial and temporal dynamic change on vegetation within local landscape is increasingly important. Recently, the advancement of technology in sensor web has shown to be promising for improving the estimation of local NPP due to its ability to provide real-time data. Still the challenge ahead to researchers is how to integrate remote sensing and the sensor web data with variability in spatial and temporal scale for better estimation of NPP at local level. The objective of this study was to demonstrate a method for acquiring local NPP estimations by integrating remote sensing and sensor web data at Gendt location.
The light use efficiency model has been found to be more appropriate for NPP estimation at local scale level in comparison to other methods and was adopted in this thesis. Spatial variability in remote sensing datasets were harmonized into common resolution based on aggregation technique while temporal variability in sensor web datasets were harmonized into daily time step based on integration and average techniques. Therefore, LUE method was used to integrate the remote sensing and sensor web data and the results were investigated by comparing them with two coarse scale MODIS standard products. The results attained from LUE model shows that high annual NPP values were obtained for cropland and grassland compared to other vegetation types. The comparison was done based on the derived daily GPP and annual NPP 2007 with MODIS product 8 days GPP (MOD17A2) and annual NPP (MOD17A3) 2006. Correlations were found at some locations with cropland and grassland for GPP while no correlation was found for annual NPP, though at some locations with grassland and cropland the values of NPP seemed to be much closer. Sources for the difference in results were identified as change in management of the parcel/size or land use types at some locations within the study area since the comparison was based on different years. The use of different light use efficiencies in the estimation of daily GPP is also considered as a source of differences. Additionally, the use of different parameters in the estimation of growth and maintenance respiration were recognized as sources. The demonstrated method was successful for data integration but the validity of the results obtained need further study.
Keywords: Gross primary production, Net primary production, Light use efficiency, MODIS, Remote sensing, In-situ sensing (sensor web).