Comparing MODIS Net Primary Production Estimates with Terrestrial National Forest Inventory Data in Austria

Mathias Neumann,Hubert Hasenauer,Maosheng Zhao,Georg Kindermann

doi:10.3390/rs70403878

Abstract

The mission of this study is to compare Net Primary Productivity (NPP) estimates using (i) forest inventory data and (ii) spatio-temporally continuous MODIS (MODerate resolution Imaging Spectroradiometer) remote sensing data for Austria. While forest inventories assess the change in forest growth based on repeated individual tree measurements (DBH, height etc.), the MODIS NPP estimates are based on ecophysiological processes such as photosynthesis, respiration and carbon allocation. We obtained repeated national forest inventory data from Austria, calculated a “ground-based” NPP estimate and compared the results with “space-based” MODIS NPP estimates using different daily climate data. The MODIS NPP estimates using local Austrian climate data exhibited better compliance with the forest inventory driven NPP estimates than the MODIS NPP predictions using global climate data sets. Stand density plays a key role in addressing the differences between MODIS driven NPP estimates versus terrestrial driven inventory NPP estimates. After addressing stand density, both results are comparable across different scales. As forest management changes stand density, these findings suggest that management issues are important in understanding the observed discrepancies between MODIS and terrestrial NPP.

Highlights

Productivity estimates are important measures to characterize the mass budget of a forest ecosystem
For the presented results and figures, the median is used which is less affected by outliers and skewed distributions
In the text the arithmetic mean is given, any comparison based on mean values assumes a symmetric distribution of the results and a balanced age class distribution of forests

Summary

Introduction

Productivity estimates are important measures to characterize the mass budget of a forest ecosystem. In this context, the carbon balance and carbon storage of the earth’s ecosystems and their spatial and temporal development is important. Large scale carbon measures are currently provided by National Forest Inventories, flux towers and remotely sensed methods, such as the MODIS Pan et al [2] observed deviations between MODIS and terrestrial NPP using inventory data from the mid-Atlantic region of the USA and suggest that differences in water availability explain this variation. Hasenauer et al [3] found similar deviations and concluded that the different data sources for predicting NPP compare well after addressing forest management impacts. The authors further suggest that a combination of “ground-based” forest data with “space-based” forest productivity estimates would utilize the advantages of both methods

Objectives

Methods

Results

Discussion

Conclusion