Skip to navigation Skip to content

Developing a Gaussian process emulator of the Nitrogen Risk Assessment Model for Scotland (NIRAMS)

Mean groundwater nitrate concentrations under different fertilizer input scenari
Mean groundwater nitrate concentrations under different fertilizer input scenarios

Diffuse pollution from agriculture is a significant problem for both surface and groundwater quality, potentially affecting ecosystems and human health.


The Nitrogen Risk Assessment Model for Scotland (Dunn et al., 2004a, 2004b) is a spatially distributed model developed and applied to quantify agricultural nitrogen exports to surface and groundwater in Scotland. Nitrate leaching is simulated for 1 km × 1 km grid cells covering the whole of Scotland based on climatic variables, inorganic fertilizer and manure inputs and land use and soil characteristics. The model includes four parameters that can be estimated by calibration describing the proportion of nitrogen in organic manure available for leaching, and the rates of mineralisation, denitrification, and leaching, respectively. The long computational times of complex process-based models such as NIRAMS limit their applicability with respect to uncertainty estimations based on multiple parameter sets and for simulations of multiple scenarios of changes in climate and land management.


In this project, a Gaussian process emulator for NIRAMS is developed which is a statistical model that mimics the result of NIRAMS in a computationally more efficient way allowing for uncertainty estimates of model results. The Gaussian process emulator has been constructed under baseline conditions (2010-2015, observed climate and land management as per the agricultural census) and for various climate and land management scenarios, using 81 parameter sets in each case. The results of the emulator include mean nitrate concentrations in individual cells and groundwater bodies and exceedance frequencies of thresholds.


References
Dunn, S.M., Lilly, A., DeGroote, J., Vinten, A.A., 2004a. Nitrogen risk assessment model for Scotland: II. Hydrological transport and model testing. Hydrol. Earth Syst. Sci. 8, 205–219. https://doi.org/10.5194/hess-8-205-2004


Dunn, S.M., Vinten, A.J.A., Lilly, A., DeGroote, J., Sutton, M.A., McGechan, M., 2004b. Nitrogen risk assessment model for Scotland: I. Nitrogen leaching. Hydrol. Earth Syst. Sci. 8, 191–204. https://doi.org/10.5194/hess-8-191-2004

 

Research

Areas of Interest


Printed from /research/developing-gaussian-process-emulator-nitrogen-risk-assessment-model-scotland-nirams on 24/10/19 06:28:46 AM

The James Hutton Research Institute is the result of the merger in April 2011 of MLURI and SCRI. This merger formed a new powerhouse for research into food, land use, and climate change.