Reconstructing and Mapping Annual Net Primary Productivity (NPP) Since 1940 Using Tree Rings in Southern Indiana, U.S.

Abstract

Vegetation absorption is one major form of carbon storage. The earliest spatial distribution of the Net Primary Production (NPP), an index to estimate how much carbon is absorbed, could extend back to the 1980s from satellite imagery. Our study reconstructed a time series annual NPP maps in the southern Indiana since 1940 with point-by-point regression models and ring-width index (RWI) from 16 tree-ring chronologies. Our RWI-NPP model had a good performance using Random Forest (RF) regression comprehensively considering both normal and dry years. The RWI-NPP model performance gap between forest and grassland is acceptable. We also found that the tendency (model of the tendency = −0.50) based on the combination of real NPP data and simulated NPP data were opposite to the one (slope = 18.70) only based on real NPP data where the extended data set could correct some bias caused by limited data. There is a huge NPP fluctuation in the recent years (2010–2013) which is highly likely to be caused by the combination of higher frequency of extreme climate events and the intensive land-use and land-cover change. We assume that most of the vegetation pixels had the same growing pattern with the plot in Morgan Monroe Flux Tower whose dominant species is ACSH (35.66%) or the plot in Hoot Woods whose dominant species is FRAM (34.41%). This study is novel in the assessment of the spatial distribution patterns of NPP since 1940. We can witness how the NPP changes within the last 70 years.