On classification, ranking, and probability estimation.

Abstract

The aim of this thesis was the study of respiration in ocean margin \nsediments and the assessments of tools needed for this purpose.\n \n \nThe first study was on the biological pump and global respiration \npatterns in the deep ocean using an empirical model based \non sediment oxygen consumption data. In this thesis the depth \ndependence of respiration patterns was modelled using a compiled data \nset of sediment oxygen consumption rates. We showed that the depth \nrelationship can best be described by a double exponential model. By \nassuming a uniform flux laterally across the global ocean the depth \nattenuation of POC could be derived. The results from this study imply \na more efficient biological pump.\n \n \nThe second study was on the short-term fate of phytodetritus, \ninvestigated across the Pakistan margin of the Arabian Sea. Stations \nranged in water depths from 140 to 1850 m, encompassing the oxygen \nminimum zone. Phytodetritus sedimentation events were simulated by \nadding 13C-labelled algal material to surface sediments. The labelled \ncarbon was subsequently traced into bacterial lipids as a proxy of \nbacterial biomass, foraminiferan and macrofaunal biomass as well as \ninto dissolved organic and inorganic pools. The largest pool of \nprocessed carbon was found in dissolved inorganic carbon, attributed \nto respiration. Macrofaunal influence was most pronounced at the lower \npart of the oxygen minimum zone. \n \n \nThe third study was on benthic respiration rates in the Gulf of \nFinland, Baltic Sea. Rates were based on in situ incubations \nusing benthic chamber landers. Three contrasting stations with \ndifferent sediment accumulation regimes were visited. The effect of \nchanges in water masses on the benthic fluxes was investigated with a \ndynamic diagenetic model. Fluxes of dissolved inorganic carbon were \nhighest at the station with accumulation bottom, intermediate at the \nstation with transport bottom and lowest at the station with erosion \nsediments.\n \n \nThe fourth study was primarily a theoretical investigation of how well \na certain model parameter can be constrained based on a certain \ndataset. A model of bio-irrigation was used as a good example of this. \nThe interpretation is based on fitting observed data with a model \ncontaining several parameters, where some parameters are a \n priori unknown. In this chapter, it was tested under what \nconditions the results obtained through this fitting are robust. The \nresults from this study imply that using only the concentration \nchange in the overlying water, it is not possible to constrain both \nthe rate and the mechanism of bio-irrigation, thus, sampling the \nporewaters at the end of the incubation is a necessity.\n \n \nThe fifth study was an an evaluation of the performance of an oxygen \noptode. The performance of the sensor was evaluated and compared with \ndata obtained by other methods. The principal conclusion was that, \nowing to high accuracy, long-term stability (more \nthan 20 months), lack of pressure hysteresis and limited \ncross-sensitivity, this method is overall more suitable for oxygen \nmonitoring in the aquatic environment than other methods.