In this thesis are shown the first results from the flight surveys measuring atmospheric carbon dioxide (CO2) in the lower troposphere (<4000 m high) in the central area of the Ebre basin following the 42°N parallel. The aim of these flights is to improve the knowledge of the carbon budget in the northeast of the Iberian Peninsula by characterizing the distribution of atmospheric CO2 in relation to carbon surface fluxes and atmospheric transport processes; and the characterization of the CO2 seasonal cycle in the boundary layer and in the free troposphere. The first flight took place in 2004, stepping up the sampling frequency at the end of 2005 and until 2007.

In order to better know the CO2 dynamics and its spatial distribution a new design of atmospheric sampling in aircrafts is presented: the Crown Atmospheric Sampling (CAS). It involves measuring the CO2 concentration along the sides and the edges of a triangular prism of air, sampling horizontally its base at 600 m above sea level (masl) and at 1200 masl; and vertically along its vertices from 600 masl up to 2500 masl of height. One of its vertices is sampled up to 4000 masl. This method measures the CO2 concentration and its variability within the boundary layer in regard to surface fluxes and atmospheric transport; and measures the concentration in the free troposphere representative for a watershed. The CAS improves the measurement of the CO2 variability compared to vertical profiles or oscillating flights. The first region where the CAS was applied was the triangle bordered by the towns of Linyola, Mequinensa and Binèfar in February 2006. Later on it was extended in the triangle of La Muela, Tarazona and Egea de los Caballeros in October 2006.

In this thesis the structure of the distribution of the CO2 concentration with regard to the intensity of vertical mixing is characterized. The variability of CO2 concentration within the boundary layer is given by the intensity of surface fluxes when crossing regions with different land use and/or photosynthetic activity. The spatial structure of CO2 in the volume sampled is given by the vertical atmospheric stability, which keeps or fades away the footprint of the surface fluxes.

A time series of atmospheric CO2 concentration both within the boundary layer and the free troposphere is built using the observations from the monthly flights in 2006; and the bimonthly ones in 2007. The amplitude of the seasonal cycle in the central part of the Ebre basin experiences a decrease in height: from 17 and 13 ppmv at 600 masl in the Linyola and La Muela regions, respectively; to 11 ppmv at 4000 masl. The decrease on the amplitude of the seasonal cycle in height is due to the vertical propagation of the signal of surface fluxes seasonal changes. There is a lag of one month in the registration of the seasonal minimum for the time series in height (>2500 m) compared to the registered at lower altitudes (600 and 1200 masl) where the minimum takes place in August. The time series of the CO2 variability contained in a horizontal scale of ~100 km provides the range of values of the representation error for atmospheric inversion models, from 0.4 up to 3.7 ppmv for Linyola's CAS; and from 0.4 ppmv to 1.3 ppmv in La Muela's CAS.

The Potential Surface Influence (PSI) area for the CO2 observations from the flight surveys has been estimated by means of the Lagrangian Particle Dispersion Model (LPDM) FLEXPART in backward mode. The PSI is the area between 0 and 300 m above the ground level (magl) where air masses potentially resided before arriving at the sampling site or also called receptor, and is expressed in units of residence time (seconds). The PSI relates the influence of surface fluxes from upwind regions to the concentration measured at the site or receptor. A new method has been proposed to assess the influence of upwind surface fluxes at the regional scale (104 km2). The Regional Potential Surface Influence (RPSI) area is composed by those grid cells in the surface layer (0-300 magl) where air masses have resided more than 500 seconds in the 4 days prior the sampling. The application of Principal Component Analysis (PCA) on all RPSI over a year for each of the vertical profiles of a CAS sets the factors which influences the most in the RPSI and divides the influence regions in local or regional; and classifies the regional depending on the weather pattern.

The use of the RPSI assesses the temporal range of the regional surface influence which is confined 57 and 21 hours prior the sampling at 600 masl and 1200 masl, respectively, in the Ebre watershed. Conversely, at 2500 masl and 4000 masl the influence of the atmospheric layer which has been in contact with the surface only occurs during spring and summer. The concentration above the boundary layer is none or little influenced by the variability of surface fluxes at the regional and local scales and thus, its measurement every ~100 km is appropriate to know the background concentration.

The scales of the PSI and RPSI explain the CO2 differences measured in sites separated by 60 km. While sites share 90% of the PSI, the common area of influence is reduced to 70% on the regional scale (RPSI). CO2 differences between profiles of the same CAS are due to either the influence of specific local fluxes on each vertex or the influence of fluxes with an origin on the unshared RPSI area. The CO2 vertical gradient is given by the physical state of the atmosphere at the time of sampling and the presence of atmospheric layers with different dynamics with regard to the transport and horizontal mixing. The study of the increase of the PSI area over time and the type of transport and mixing of air masses characterize and identify consistent atmospheric layers which are not dispersed during transport. Thus, they preserve the signal of fluxes from the areas that have swept. The dynamics of air masses horizontal transport before reaching the sampling site is characterized by the dispersive exponent (¿), the Lyapunov exponent (¿) and the exponent that characterizes the increase of the PSI area over time (ß). The transport of air masses reaching the lower troposphere (<4000 m) in the Ebre basin is superdifussive (¿>1/2) and the horizontal mixing follows a stochastic dynamics (potential Lyapunov exponent ¿>0) for all seasons in 2006. The vertical profiles of these indices indicate that there is a discontinuity in the atmosphere in terms of horizontal transport, located at 2500 m on La Muela's profile. Above it, transport is laminar, while below it, chaotic mixing and filamentation dominate the dynamics of horizontal mixing. The vertical profiles of these indices are proposed as an alternative method for identifying atmospheric layers containing the carbon signal of upwind regions influencing the vertical distribution of CO2 in the lower troposphere.

Sampling vertical profiles and horizontal transects at different times during a day allows the estimation of the average surface flux in the central area of the Ebre watershed by means of budget integrative methods such as the Integrative Convective Boundary Layer and the Carbon Content Integration. These methods use the changes in the CO2 concentration in the air column and the changes in the boundary layer height to estimate the average surface flux. The central part of the Ebre basin uptakes atmospheric CO2 with a flux of (-2.4) - (-7.9) ¿molCO2¿m-2¿s-1 in a surface of 2¿105 km2 on the 26th June 2007 between 8 and 18 UT.

When extrapolating this flux into annual values, the central part of the Ebre basin appears to be a weak sink of atmospheric CO2 uptaking an annually net amount of CO2 of -0.2GtC.yr-1. European ecosystems are estimated to have a mean surface flux from -0.9 to +0.2 GtC.yr-1. The Ebre basin represents a 0.8% of the total European continent surface and then, its contribution to the integral budget of atmospheric carbon sources and sinks is small.

The tropospheric flights in the Ebre basin complement the network of stations sampling both the boundary layer and the free troposphere on the European continent. The data from these flights would improve the estimates of surface fluxes in South Europe from atmospheric inversion models and would be use for satellite and FTIR data validation.

Font Font, Anna Maria

Xavier Rodó López i Tutora: Ileana Bladé Mendoza

Thursday, 15 December, 2011

SALA DE GRAUS (AULARI)

Jaume Piera Fernández, Vicedirector i Cientific Titular.Centre Mediterrani d'Investigacions Marines i Ambientals (CSIC). (President).
Jan Polcher. Director de Recerca. Institut Català de Ciències del Clima. (Secretari).
Laszlo Haszpra. Principal senior councilor at the Hungarian Meteorological Service. Orszagos Meteorologiai Szolgalat in Hungarian. (Vocal).

11:00