MyLake

General Information
Acronym of the model	MyLake
Full name of the model	Multi-year simulation model for Lake thermo- and phytoplankton dynamics
Model components	Hydrodynamics, Chemistry, Biology
Supported platforms	Windows, Mac, Linux
Programming Language	MATLAB
Still maintained	Yes, by: Raoul-Marie Couture (et al.)
Most recent version	MyLake v1.2 (public); MyLake v2.0 (+ sediment) & MyLake-C (+ CH4)
Model structure
1D
Other: Flexible grid
Model description
Model objective	Investigate, predict, or test hypotheses related to processes controlling ice cover duration, lake hydrodynamics, and sediment-water interactions
Specific application	“The MyLake lake model is best suited for applications having the following characteristics: 1) the geometry of the lake basin lends itself to the 1D assumption, which neglects lateral heterogeneity, or if you can accept the limitation of working with a single 1D profile; 2) you favour short integration time over model complexity; 3) you do not need complex lake physics or ecological modelling (e.g., saline or groundwater intrusions, reservoir management, food-web, etc.); 4) you want access to the source code and would like to modify it yourself if necessary; 5) the lake experiences seasonal ice cover.” (Couture & Tominaga 2016)
Background knowledge needed to run model	Basic knowledge of MATLAB programming
Basic procedures	For one model time step(24 h): 1) Calculate surface heat fluxes and wind stress, light attenuation, and phytoplankton growth and loss rates 2) Calculate profile of the diffusion coefficient K 3) Calculate the heat flux between water and sediment (taken as heat source/sink for each layer) 4) Solve new profile for each state variable taking into account advection, diffusion and local sources/sinks. Solving is done in following order: 1) temperature, 2) tracers, 3) chlorophyll-a, 4) particulate phosphorus, 5) dissolved phosphorus. 5) Update content of the sediment stores of chlorophyll a and particulate phosphorus A) If no ice: i) Check for (and possibly proceed) autumn/spring turnover B) If ice cover: a) If Ta<Tf (freezing) i) Calculate ice surface temperature (depending on snow cover, or ice thickness if snow is absent) ii) Calculate snow ice formation in case of isostatic imbalance iii) Calculate congelation ice growth by Stefan’s law iv) Accumulate new snow fall and subtract formed snow ice from snow cover b) If Ta≥Tf (melting) i) Melt snow or ice from top with total surface heat flux C) Melt ice from bottom with the heat diffused to the surface layer (keeping the surface layer temperature at freezing point) D) Update snow density 6) Add river inflow and update profiles of the state variables accordingly 7) Mix unstable layers until stable density profile 8) Mix water layers with the available turbulent kinetic energy TKE (TKE=0 under ice cover) 9) Check for supercooled layers and turn them into (initial) ice 10) Save results to output matrices (Return to Start) (Saloranta & Andersen 2007, Markelov et al. 2019)
MyLake (Multi-year simulation model for Lake thermo- and phytoplankton dynamics) is a one-dimensional process-based lake model coded in Matlab whose initial release was developed in 2007 by Tuomo Saloranta and Tom Andersen at the Norwegian Institute of Water Research. It has a relatively simple and transparent model structure, it is easy to set up, and is suitable both for making predictions and scenarios, and to be used as an investigative tool. Short runtime allows application of comprehensive sensitivity and uncertainty analysis as well as simulation of a large number of lakes or over long periods (decades). MyLake aims to include only the most significant physical, chemical and biological processes in a well-balanced and robust way (Saloranta & Andersen 2007). The model is available as an open-source project via GitHub. Published versions are made public and unpublished version accessible via password. As of Fall 2019, three institutions now contribute to the code: i) the Norwegian Institute for Water Research, Norway, ii) Laval University, Canada, and ii) the University of Helsinki, Finland. Being an open source project, the MyLake code has been forked and adapted to suit a project’s needs. There are thus several existing branches, which are collected in a suite of repositories at github.com/biogeochemistry. Here we identify three versions of MyLake that are actively used or developed: i) 1.12 refers to the Mylake model as it was first written by Saloranta and Andersen (2007). MyLake 1.12 is for the simulation of daily 1) vertical distribution of lake water temperature and thus stratification, 2) evolution of seasonal lake ice and snow cover, and 3) phosphorus-phytoplankton dynamics. It contains a simple sediment box-model. When only hydrodynamic and ice cover simulations are necessary, this version is recommended with sediment and phytoplankton module turned off. ii) Version 2.0 refers to the model with extended biogeochemical reaction network ( N, O, P, Mn, Fe, Al, Ca, Si, S, pH) and process-based sediment diagenesis with a focus on redox processes. It builds on version 1.12 and adds simulation of daily 1) distribution of geochemical species, 2) fluxes across the sediment-water interface, 3) vertical distribution of geochemical species in the sediment and porewater. For simulation focussing on oxygen dynamics, either the diagenetic module or a simpler fixed-k sediment oxygen demand module can be used. iii) Version C refers to the model with extended gas exchange models for O, CO2 and CH4 with a focus on DOC cycling. It builds on version 1.12 and adds simulation of daily distribution of dissolved cases according to user-selectable gas exchange models. Links to the source code and manuals for those versions are given below.
Link to website/manual	Version 1.12 Code “Manual”: https://niva.brage.unit.no/niva-xmlui/handle/11250/212445 Version 2 : Code Manual: Under review Version C : Code Manual
Model characteristics
Input variables	Obligatory: – Lake morphometry and initial profiles - Meteorological forcing (air temperature, air pressure, wind speed, relative humidity, precipitation, irradiance) Optional: - Inflow loadings (water volume, particulates, dissolved species)
Input file format	.xls Other, namely: MyLake v2.0 and vC requires .txt
Output variables	Added in v1.12 • Temperature profile • Snow and ice cover (presence/absence, thickness, temperature) • Chlorophyll-a profile • Light attenuation profile • Sediment heat flux Added in v2 • Dissolved oxygen and carbon dioxide • Solid-phase and dissolved constituents • Sediment and porewater concentration profiles Added in vC • Dissolved oxygen, carbon dioxide • Carbonate equilibria
Output file format	Other, namely: .mat (can be exported from MATLAB as .csv or other file format)
Biogeochemical model components	• Nutrients: inorganic P, particulate P, dissolved nitrate • Phytoplankton: phytoplankton biomass C, measured as chlorophyll-a and dissolved inorganic phosphorus (phosphate) P; fixed C:N:P composition • Zooplankton: not included in model • Geochemistry: N, Mn, Fe, Al, Ca, S, Si
Model structure/mathematical framework	Partial differential equations
Temporal resolution	Daily
Minimal spatial resolution	0.5 m
Variables needing calibration	– Primary: water temperature (dissolved oxygen in v2.0) - Additional: nutrients, chlorophyll-a
Has successfully been used in
Climate Change Scenario	Couture et al. 2014 Couture et al. 2015 Kiuru et al. 2018
Oligotrophic water	Pilla et al.
Mesotrophic water	Kiuru et al. 2019 Kiuru et al. 2018
Eutrophic water	Markelov et al. Couture et al. 2018 Couture et al. 2014
Management Support	Moe et al. 2019 Moe et al. 2016
Which institutes have applied the model	Norwegian Institute for Water Research (NIVA) Norwegian Technical University (NTNU) University of Oslo Finnish Environmental Institute (SYKE) University of Helsinki Swedish Agricultural University (SLU) University of Waterloo Laval University University of Québec INRS-ETE
Has coding for	Ice dynamics, sediment heat flux, sediment dynamics, phytoplankton biomass
Accessibility
Open-Source, Open-to-use, Test cases available
Available tools for pre- and post-processing	Some exemplary tools from individual applications available at https://github.com/biogeochemistry Users have made available tools in Python, R and Matlab for optimisation and post-processing
Support	GitHub “issues” and “wiki” at https://github.com/biogeochemistry
Can be coupled to the following models	INCA-Phosphorus, INCA-Carbon, PERSiST Simply-P
How can someone get access to this model	To gain access to the Mylake private GitHub send your github username to: raoul.couture@chm.ulaval.ca
Miscellaneous
Comments	Development of new versions and revisions often occur in parallel due to the open source nature of the model. The MyLake GitHub keeps these rather up-to-date with relevant applications, publications, and manuals for new versions.
Useful tricks and hints	Publicly-available application of MyLake in R
Links	Manual MyLake GitHub Toolbox
Form was updated: 2019-09-06

Reference list:

Saloranta & Andersen 2007
“Gebre et al. 2014”: doi:10.5194/tc-8-1589-2014
Kiuru et al. 2019
Kiuru et al. 2018
Couture et al. 2018
Couture et al. 2014
Moe et al. 2019
Moe et al. 2016