Message Boards

BLOOM: conversion fractions in mineralization process

Francesco P, modified 5 Months ago.

BLOOM: conversion fractions in mineralization process

Youngling Posts: 1 Join Date: 1/23/19 Recent Posts

Hi all,


I'm new to this forum but it's been a while now since I started working with Delft3D FLOW/BLOOM. Our main objective is to simulate the strong algal (mainly cyanobacterial) blooms occurring in a shallow pond that has no inlets nor outlets. The main issue with our present configuration is the lack in recycling of nutrients, and nitrogen in particular. The model gives satisfying results during the growth phase, but after the pick of growth has been reached concentrations of NH4 and NO3 are not strong enough to sustain growth and the population slowly decreases.


We are therefore willing to perform a calibration of the parameters included in our .inp file (see attachment) through Bayesian methods and automatic multiple runs on a short period (around two weeks).


Along the serial mineralization from POC1 to POC4, parameters like b_poc1poc2 and b_poc1doc are described on the technical reference manual as "conversion fractions" and it is suggested to set them, respectively, to 1.0 and 0.025.

Does this mean that their sum can be higher then 1 without interfering with the general mass balance? During our random calibration should we force their sum to always be equal to 1 ?


I attach our .inp file, as well as a few plots on the evolution of cyanobacteria and nutrients concentration at one location, along time and depth.


Thanks for your help,