Resumen

MOLINA-CASTRO, Gabriel. Evaluation of correction factors to estimate the uncertainties of triangular distributions with 95% coverage intervals. Ingeniería [online]. 2022, vol.32, n.2, pp.15-31. ISSN 2215-2652.  http://dx.doi.org/10.15517/ri.v32i2.49699.

Recently, uncertainty estimation has become a metrological requirement of great interest for the reporting of greenhouse gas inventories in Costa Rica. In this context, a methodological guide published by the National Carbon Neutrality Program has emerged as the basis for the local development of this issue. However, additional efforts are still pending to provide a clear technical explanation that justifies the implementation of some aspects of its content. The present study assesses the validity of the use of correction factors (FC), proposed by this guide, that enlarges the standard uncertainty estimated from a triangular distribution, in the presence of a 95 % coverage interval. To achieve this, 3124 triangular distribution scenarios were simulated using R statistical software. Two standard uncertainties were estimated for each scenario: one considering that the simulated limits delimit a 100 % coverage interval (u100) and another considering a 95 % coverage interval (u95), with two possible interpretations about the location of the interval for the latter. FC were estimated for four groups as the mean ratio between u95 and u100. Results between 1.20 and 1.29 were obtained for FC, with a global value of 1.25 and no significant differences between them. These values are highly consistent with those recommended in the national methodological guide, thus verifying their validity and applicability. Finally, a unique FC equal to 1.25 is suggested as a simple practical approximation for all the evaluated scenarios, easing its implementation by users.Keywords:

Palabras clave : Greenhouse gases; National Program for Carbon Neutrality; probability distribution; simulation; triangular distribution; uncertainty.

        · resumen en Español     · texto en Español     · Español ( pdf )