Status: this preprint was under review for the journal BG but the revision was not accepted.
Inferring particle dynamics in the Mediterranean Sea from in-situ
pump POC and chloropigment data using Bayesian statistics
Weilei Wang,Cindy Lee,and Francois Primeau
Abstract. Chloropigment and particulate organic carbon (POC) concentration data collected using in-situ large-volume pumps during the MedFlux project in the Mediterranean Sea in May 2005 provided an opportunity to estimate rate constants that control the fate of particles and specifically chloropigments in the water column. Additionally, comparisons to thorium and chloropigment data from settling-velocity (SV) sediment traps at the same site enabled us to distinguish between the influence of the sampling method used vs. the tracer used on particle dynamic rate constants. Here we introduce a Bayesian statistical inversion method that combines the data with a new box model and has the capacity to infer rate constants for POC respiration/dissolution, chlorophyll and phaeopigment degradation, and particle aggregation and disaggregation. The estimated small-particle (1–70 μm) POC respiration rate constant was 1.25+0.55−0.38 yr−1 (0.80 yr). For this data set, the rate constants for chlorophyll (Chl) degradation to phaeopigments and phaeopigment respiration were not well constrained. The estimated aggregation and disaggregation rate constants were 7.65+3.35−2.33 (0.13 yr) and 106.09+39.13−28.59 yr−1 (0.01 yr), respectively, which indicates that particle aggregation and disaggregation were extensive at the studied depths (125–750 m) in May after the spring bloom had ended and flux was low.
How to cite. Wang, W., Lee, C., and Primeau, F.: Inferring particle dynamics in the Mediterranean Sea from in-situ
pump POC and chloropigment data using Bayesian statistics, Biogeosciences Discuss. [preprint], https://doi.org/10.5194/bg-2018-6, 2018.
Received: 05 Jan 2018 – Discussion started: 13 Mar 2018
We introduce a new model that can infer particle dynamics rate constants using large volume pump chloropigment data. We compare the inferred rate constants to those estimated using the same tracers but collected using sinking velocity (SV) sediment traps, and to those estimated using SV trap thorium data. We conclude that the two different sampling methods have less influence on particle aggregation and disaggregation rate constant estimations than do the two different tracers themselves.
We introduce a new model that can infer particle dynamics rate constants using large volume pump...