CEM - Abstract

Abstract Information :

Agriculture is a major contributor to global gaseous emissions, particularly greenhouse gases (GHGs) such as CO₂, CH₄, and N₂O, as well as ammonia (NH₃), which impacts air quality. These GHGs represent 10% of the total emission worldwide, and 95% for ammonia. Consequently, accurate field measurements on livestock buildings which are an important source is crucial in order for reliable emission reductions from these facilities, which are essential to drive effective mitigation strategies. The most commonly used methods rely on quantifying the air exchange rate and the concentration gradients of the target gases. In naturally ventilated buildings such as barns, the quantification of the air exchange rate is based on the CO₂ produced inside the building and considered as an internal tracer. In the barn CO2 is produced by animal and manure (and heating systems eventually). When analyzing the model used to quantify emissions, it is clear that the accuracy of this method depends strongly on estimating CO2 production by the animals. As part of the quantiAGREMI project, one key objective was to address uncertainty in field measurements of gaseous emissions from livestock buildings, such as CO2 emission calculated from heat production models. The CO₂ emissions from different animal categories can be quantified based on heat production models available in the literature. In this study, we quantified the associated uncertainties using existing datasets for dairy cows, fattening pigs, and broiler chickens. We assessed the uncertainty on CO₂ emissions from livestock using both the GUM method (Guide to the Expression of Uncertainty in Measurement) based on analytical propagation, and the Monte Carlo method as described in GUM Supplement 1, based on propagation distribution. Two types of results were obtained, based on three different animal datasets. For dairy cows, the calculations are independent of time. In contrast, for fattening pigs and broilers, the estimates depend on time due to animal growth because heat production of these animals depends on body mass, which evolves over time. To estimate the growth, Gompertz equations were fitted to a few recorded body weights. As a result, the main source uncertainty of heat production for pigs and broilers is body mass. For dairy cows, heat production was estimated from the body mass, fat-corrected milk yield, and pregnancy stage, and it turned out that in this case, that was the pregnancy and fat-corrected milk which represented the main source of uncertainty. Overall, uncertainty estimates obtained with the GUM and Monte Carlo methods were similar. For the simulations with the experimental datasets, the uncertainty of heat production was very low whatever the animal category because animal body weights were measured accurately. As conclusion, heat production estimates can be a robust basis for estimating CO₂ emissions from livestock.