Development of new historical global Nitrogen fertilizer map and the evaluation of their impacts on terrestrial N cycling and the evaluation of their impacts on terrestrial N cycling

Abstract:

The use of synthetic nitrogen fertilizer was rapidly growing up after the birth of Haber-Bosch process in the early 20th century. The recent N loading derived from these sources on terrestrial ecosystems was estimated 2 times higher than biogenic N fixation in terrestrial ecosystems (Gruber et al., 2009). However, there are still large uncertainties in cumulative N impacts on terrestrial impact at global scale. In this study, to assess historical N impacts at global scale, we made a new global N fertilizer input map, which was a spatial-temporal explicit map (during 1960-2010) and considered the fraction of NH₄⁺ and NO₃^- in the N fertilizer inputs. With the developed N fertilizer map, we evaluated historical N20 cycling changes by land-use changes and N depositions in N cycling using ecosystem model ‘VISIT’.

Prior to the downscaling processes for global N fertilizer map, we applied the statistical data imputation to FAOSTAT data due to there existing many missing data especially in developing countries. For the data imputation, we used multiple data imputation method proposed by Honaker & King (2010). The statistics of various types of synthetic fertilizer consumption are available in FAOSTAT, which can be sorted by the content of NH₄⁺ and NO₃^-, respectively. To downscaling the country by country N fertilizer consumptions data to the 0.5˚x 0.5˚ grid-based map, we used historical land-use map in Earthstat (Rumankutty et al., 1999). Before the assignment of N fertilizer in each grid, we weighted the double cropping regions to be more N fertilizer input on to these regions. Using M3-Crops Data (Monfreda et al., 2008), we picked up the dominant cropping species in each grid cell. After that, we used Crop Calendar in SAGE dataset (Sacks et al., 2010) and determined schedule of N fertilizer input in each grid cell using dominant crop calendar. Base fertilizer was set to be 7 days before transplanting and second fertilizer to be 30 days after base fertilizer application.

First, we compared the differences in historical changes of geographic distribution between the new fertilizer map and N deposition maps. Finally, we introduce how the accounting NH₄⁺:NO₃^- ratio in global fertilizer changes the amount of global N₂O emission in ecosystem model simulations using VISIT.