Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Japon → Characteristics of Scalar Admixture in the Atmospheric Boundary Layer and Estimation of Regional Surface Fluxes Over Semi-arid Area

University of Tsukuba (2006)

Characteristics of Scalar Admixture in the Atmospheric Boundary Layer and Estimation of Regional Surface Fluxes Over Semi-arid Area

KOTANI Ayumi

Titre : Characteristics of Scalar Admixture in the Atmospheric Boundary Layer and Estimation of Regional Surface Fluxes Over Semi-arid Area

Auteur : KOTANI Ayumi

Université de soutenance : University of Tsukuba

Grade : Doctor of Philosophy in Science 2006

Résumé
Turbulence data obtained by aircraft observations in the mixed layer was analysed to estimate the regional surface heat fluxes through the application of the variance methods. Several heights within and above the mixed layer were flown repeatedly above the flux observation site in a steppe region in Mongolia. The observed profiles of temperature and humidity showed vertical development of the mixed layer and scatter of the second order moments, i.e., the variance, near the top due to the effects of entrainment heat flux from the above atmosphere. The vertical profiles of the dimensionless temperature variance were found to follow, in most of the cases, the functional forms proposed in previous studies. These variance statistics were applied to the variance formulations to estimate surface sensible heat fluxes. First, the flux estimation was made with these equations and the constant parameters as derived in previous studies. Then, the constants were re-calibrated with the current data set and eddy correlation measurements on the ground. These constants were, then, used for the second flux estimation. Finally, additional variables, which represent the large scale atmospheric conditions, namely baroclinicity and horizontal temperature advection, were considered for improvement of the flux estimation. The resulting root mean square difference of the sensible heat flux by estimation and ground based measurements was reduced from about 40-100 W m-2 for the results obtained with the constants and formulations by the previous studies, to 30 W m-2 or less for those obtained with locally calibrated constants and introduction of four additional variables. Difference among types of formulation was not significant, and thus the usage of simpler formulation, which needs less number of parameters, is preferable for the practical application of the variance methods. The major cause of estimation error in variance methods was error of temperature variance, which is possible to be maximal at the higher level for free convection formulation and middle level for the others. The present analyses reveal that the mixed layer variance methods are capable of producing surface fluxes with turbulence data measured from an aircraft. However, it also indicates that there remains some uncertainty, which partly comes from the sampling error of temperature fluctuations by aircraft observation due to insufficient data length and the reliability in evaluation of the surface flux as reference value and its spatial representation. As a whole, it is not clear at this point whether or not the need of the local calibration is an indication of the lack of universality of the equations, given the wide range of data sets employed in the past. The need for calibration means that these experimental constants would contain uncertainties of data set and the formulations including unsuitable scaling and parameters, and the reduction of the error by the local calibration suggests that refinement of the formulations is still needed for the estimation of the surface flux by the mixed layer variance methods with sufficient accuracy.

Présentation

Version intégrale (2,9 Mb)

Page publiée le 11 janvier 2012, mise à jour le 21 avril 2020