Analysis of carbon dioxide concentration skewness at a rural site

Trend analysis and outlier distribution of CO2 and CH4: A case study at a rural site in northern Spain

CO₂ and CH₄ outliers may have a noticeable impact on the trend of both gases. Nine years of measurements since 2010 recorded at a rural site in northern Spain were used to investigate these outliers. Their influence on the trend was presented and two limits were established. No more than 23.5% of outliers should be excluded from the measurement series in order to obtain representative trends, which were 2.349 ± 0.012 ppm year^-1 for CO₂ and 0.00879 ± 0.00004 ppm year^-1 for CH₄. Two types of outliers were distinguished. Those above the trend line and the rest below the trend line. Outliers were described by skewed distributions where the Weibull distribution figures prominently in most cases. A qualitative procedure was presented to exclude the worst fits, although five statistics were considered to select the best fit. In this case, the modified Nash-Sutcliffe efficiency is prominent. Finally, three symmetrical distributions were added to fit the observations when outliers are excluded, with the Gaussian and beta distributions providing the best fits. As a result, certain skewed functions, such as the lognormal distribution, whose use is frequent for air pollutants, could be questioned in certain applications.

Influence of Wind Speed on CO2 and CH4 Concentrations at a Rural Site

Meteorological variables have a noticeable impact on pollutant concentrations. Among these variables, wind speed is typically measured, although research into how pollutants respond to it can be improved. This study considers nine years of hourly CO₂ and CH₄ measurements at a rural site, where wind speed values were calculated by the METEX model. Nine wind speed intervals are proposed where concentrations, distribution functions, and daily as well as annual cycles are calculated. Contrasts between local and transported concentrations are around 5 and 0.03 ppm for CO₂ and CH₄, respectively. Seven skewed distributions are applied, and five efficiency criteria are considered to test the goodness of fit, with the modified Nash-Sutcliffe efficiency proving to be the most sensitive statistic. The Gumbel distribution is seen to be the most suitable for CO₂, whereas the Weibull distribution is chosen for CH₄, with the exponential function being the worst. Finally, daily and annual cycles are analysed, where a gradual decrease in amplitude is observed, particularly for the daily cycle. Parametric and nonparametric procedures are used to fit both cycles. The latter gave the best fits, with the agreement being higher for the daily cycle, where evolution is smoother than for the annual cycle.

CO2 spatio-temporal analysis in the Iberian Peninsula

A comparison between monthly CO₂ values calculated in the Iberian Peninsula and those measured during six years commencing on October 2010 in the centre of its upper plateau is presented. Gaussian and Epanechnikov kernels are used to calculate CO₂ concentration and its growth rate in the study region from values at certain grid points. Slight spatial differences are obtained, revealing that both concentration and growth rate are nearly uniform in the region. However, some intervals were proposed that were represented by bands (strips), distributed meridionally for concentration and zonally for growth rate. Band borders were smoother for the Gaussian kernel than for the Epanechnikov kernel. In addition, the probability distribution function of concentration and growth rate were obtained with both kernels. Temporal analysis is carried out adding a linear evolution for growth rate and a sinusoidal function for the annual cycle. This revealed similar patterns for the region and at the grid point nearest to the measurement site, given by a sinusoidal function with nearly constant amplitude, providing satisfactory agreement. However, measurements showed great dispersion, with the concentration being around 7 ppm higher than for the region. Temporal evolution is characterised by a growth rate of 2.39 ppm yr^-1 and a sinusoidal function with an amplitude decrease of 0.25 ppm yr^-1.

The influence of meteorological variables on CO2 and CH4 trends recorded at a semi-natural station

CO₂ and CH₄ evolution is usually linked with sources, sinks and their changes. However, this study highlights the role of meteorological variables. It aims to quantify their contribution to the trend of these greenhouse gases and to determine which contribute most. Six years of measurements at a semi-natural site in northern Spain were considered. Three sections are established: the first focuses on monthly deciles, the second explores the relationship between pairs of meteorological variables, and the third investigates the relationship between meteorological variables and changes in CO₂ and CH₄. In the first section, monthly outliers were more marked for CO₂ than for CH₄. The evolution of monthly deciles was fitted to three simple expressions, linear, quadratic and exponential. The linear and exponential are similar, whereas the quadratic evolution is the most flexible since it provided a variable rate of concentration change and a better fit. With this last evolution, a decrease in the change rate was observed for low CO₂ deciles, whereas an increasing change rate prevailed for the rest and was more accentuated for CH₄. In the second section, meteorological variables were provided by a trajectory model. Backward trajectories from 1-day prior to reaching the measurement site were used to calculate distance and direction averages as well as the recirculation factor. Terciles of these variables were determined in order to establish three intervals with low, medium and high values. These intervals were used to classify the variables following their interval widths and skewnesses. The best correlation between pairs of meteorological variables was observed for the average distance, in particular with horizontal wind speed. Sinusoidal relationships with the average direction were obtained for average distance and for vertical wind speed. Finally, in the third section, the quadratic evolution was considered in each interval of all the meteorological variables. As regards the main result, the greatest increases were obtained for high potential temperature for both gases followed by low and medium boundary layer height for CO₂ and CH₄, respectively. Combining both meteorological variables provided increases of 22 ± 9 and 0.070 ± 0.019 ppm for CO₂ and CH₄, respectively, although the number of observations affected is small, around 7%.

Temporal patterns of CO2 and CH4 in a rural area in northern Spain described by a harmonic equation over 2010-2016

The present paper seeks to improve our knowledge concerning the evolution of CO₂ and CH₄ in terms of monthly trends, growth rate and seasonal variations in the lower atmosphere. Dry continuous measurements of CO₂ and the CH₄ mixing ratio were carried out over five and a half years (from 15 October 2010 to 29 February 2016) by multi-point sampling at 1.8, 3.7 and 8.3m, using a Picarro analyzer at a rural site in the Low Atmosphere Research Centre (CIBA), on the upper Spanish plateau. Data were divided into diurnal and nocturnal records. The mathematical equation proposed to analyze the overall data was a harmonic one, comprising a polynomial (trend) and a series of harmonics (seasonal cycle). Amplitude was considered as a constant and variable term over time. Quite different behaviour was found between day and night measurements in both climate forcing agents. CO₂ showed an accelerating trend in autumn, whereas CH₄ trends were higher during the winter. Increasing growth rates were reported for CO₂ and CH₄ over the whole study period. Nocturnal CO₂ amplitudes are higher than diurnal ones except in winter for both gases, and also in the autumn for CH₄.