Assessing the Sensitivity and Uncertainty of an NH3 Emission Reduction Calculator for Dairy Cattle Barns by Means of Monte Carlo Analysis Combined with Least Square Linearization

Mendes, L.B., Demeyer, P., Brusselman, E., & Pieters, J. (2015). Assessing the Sensitivity and Uncertainty of an NH3 Emission Reduction Calculator for Dairy Cattle Barns by Means of Monte Carlo Analysis Combined with Least Square Linearization. In: Systems Analysis 2015 - A Conference in Celebration of Howard Raiffa, 11 -13 November, 2015, Laxenburg, Austria.

Preview

Text SA2015-Assessing_the_Sensitity_of__an_NH3_Emission_Reduction.pdf - Published Version Available under License Creative Commons Attribution. Download (547kB) | Preview

Abstract

With regard to Natura 2000, the Flemish government (Belgium) established the Programmatic Approach to Nitrogen (PAS: acronym in Flemish), with the aim of reducing environmental overload of nitrogen compounds. This approach will have substantial consequences for livestock farms located next to or within special areas of conservation and will likely result in generic measures to reduce ammonia (NH3) emissions from livestock facilities. An NH3 emission reduction calculator for dairy cattle systems (AEREC-DC) was adapted based on a mechanistic approach. Reduction coefficients estimated with this tool are used to assess the efficiency of “low NH3 emission” techniques which can be implemented in Flanders at a later stage. Field measurements will be made in the future to confirm/correct them. Emission reduction techniques combining processes such as floor scraping, flushing, manure acidification, and different types of floor were modeled. The tool comprises 36 input variables, some of which have values that are based on experimental measurements. Nevertheless, reliable information concerning other relevant variables are scarce in the literature. Hence, model sensitivity analysis is imperative. We hypothesize that the ranking of input variables in terms of their effect on the model outcome will change if different uncertainty ranges are assigned to them. Hence, this study was conducted to combine Monte Carlo Analysis associated with Least Square Linearization in order to perform sensitivity and uncertainty analyses on AEREC-DC. The sensitivity analysis was performed by assigning each input variables’ probability distribution function (PDF) with a relatively narrow variance (1% of mean value). The uncertainty analysis was carried out by gradually increasing the PDF’s variance up to what is considered realistic. The outcomes of this study will help deciding which variables urgently need to be monitored experimentally in order to improve predictions’ accuracy.

Actions (login required)

View Item