Prolonged sampling time increases correlation between wind tunnel and integrated horizontal flux method.

Highlights • Urea (UR), buffalo manure (BRM) and liquid digestate (BLD) were applied on bare soil. • Open field trials were carried out during summer season under Mediterranean climate. • Two measuring methods were tested: Integrated Horizontal Flux and Wind Tunnels. • WT measurements differ from IHF results by 32% (UR), 43% (BRM) and 46% (BLD). • Prolonging sampling time exposure increases correlation between WT and IHF methods. Abstract Agriculture is still the main source of NH 3 emissions. Consequently, ammonia reduction and monitoring are included in the strict EU targets. Unfortunately, the main issue with the quantification of NH 3 emissions is that some uncertainties are still related to NH 3 volatilization measurements, mainly because of the absence of a standard measurement technique. This study compares two different methods for measuring NH 3 fluxes: the Integrated Horizontal Flux (IHF) method used with glass tubes and the wind tunnels (WT) with acid traps, in order to provide further insights into quantification of NH 3 emissions. Specifically, three field trials were carried out, using different fertilizers on bare soil: urea, raw manure and liquid separated digestate from Mediterranean buffalo cows (Bubalus bubalis L.). Results show that the same diurnal flux variations for both methods was observed, except for those in urea trials, mainly because of the dynamics of urea volatilization. Generally, WT underestimated ammonia emissions, especially during the first six hours of experimentation, since the fixed air speed inside the tunnel is lower than the outside wind speed at the same height. Cumulative ammonia volatilization measured with WT differs from those measured with IHF by 32% (urea), 43% (raw manure) and 46% (liquid separated digestate), respectively. Moreover, WT estimations seemed to be affected by a short exposition time of the acid traps, demonstrating a poor correlation of IHF and WT method within 2–3 h. Indeed, linear correlation between two methods was improved to R2 = 0.92, by increasing the minimum samplers exposition time up to 3 h. [ABSTRACT FROM AUTHOR]