幸运飞艇计划

Project description
An important aspect of climate change is the transfer of heat from the shallow to the deep part of the ocean. However, this process is not well understood. In particular, the time and spatial scales of this heat transfer have not yet been reliably characterized. A key challenge is to determine the ocean鈥檚 temperature as a function of time and position with high resolution. In this master project acoustic waves will be used to tackle this problem.

Acoustic waves systematically sample large parts of the ocean between a source and receiver. The propagation of these waves depends on the ocean鈥檚 sound velocity. As this sound velocity in its turn depends on temperature, salinity and pressure, it is, in principle, possible to determine the ocean鈥檚 temperature as a function of time and position using acoustic waves.

In order to this acoustic waves have to be modelled and then this modeling has to be used to invert observed acoustic data for temperature. The main focus in this study therefore will be on the modeling of acoustic waves. If time allows, the modelled data will be compared to observed acoustic data. The main region of interest is the north Atlantic as considerable changes in climate have been occurring in this region.

There are two main methods to model the propagation of acoustic waves through the ocean. These can be characterized as either fully numerical methods, such as finite differences, or more approximate methods based on asymptotic and perturbation expansions of the acoustic wave equation. The former are attractive, but, computationally quite expensive. In fact, for the frequencies (>100 Hz), spatial resolution (about 10 m) and longer distances (>100 km) relevant for monitoring temperature changes in the ocean, it is almost impossible to use these methods. On the other hand, the asymptotic methods, which are based on ray theory, and perturbation methods, based on single or double scattering, are approximate but relatively fast. Moreover, it has been shown that in many instances, for the
parameters of interest, these methods are quite accurate. They are therefore the method of choice in this master project.

The project will use existing computer software for modeling ocean acoustic waves using asymptotic and perturbation methods. This software will be optimized by running it on GPUs. A number of test models, representative for the north Atlantic, incorporating both small, medium and large scale structures will be developed and used in the acoustic modeling. These model initially will be purely synthetic (based on 1D velocity profiles with velocity perturbations on various scales). After that realistic ocean circulation models will be used, possibly combined with models for turbulent flow. The effect of temperature variations on the resulting synthetic acoustic waveforms, especially the travel times, amplitudes and shape of wave packages, will be studied in detail. The results of this will then be used either in setting up a synthetic inversion or in studying measured ocean acoustic data.

Proposed course plan during the master's degree (60 ECTS)
Geov112, Geov277, Geov355, Geov265, Geov232, Geov325 (or Geof338)

听