Finnish Climate Modellers’ Seminar 2020
9th Finnish Climate Modellers’ seminar - Focus on Model ensembles and intercomparisons on November 20, 2020
Registration
The seminar will be organised online on Zoom. The participation link will be emailed to the registered participants. Please register latest on November 16 via link: Registration
Schedule November 20th
Session 1
09:00 – 09:10
Welcome words
09:10 – 09:55
Dr. Juha Lento, CSC: Clean code (20 min + 10 min questions + 15 min group discussion)
10:00 – 10:30
MSc. Kalle Nordling, FMI: Using CMIP6 data from Google Cloud (20 min + 10 min questions)
10:30 – 10:45
Coffee break
Session 2
10:45 – 11:30
Keynote talk, Dr. Laura Suarez-Gutierrez, The Max Planck Institute for Meteorology: The Max Planck institute grand ensemble: Enabling the exploration of climate system variability (30 min + 15 min questions)
11:30 – 11:45
Dr. Letizia Tedesco, Finnish Environment Institute: Arctic sea ice as represented by an ensemble of CMIP5/6: findings and caveats
11:45 – 12:00
Dr. Anton Laakso, FMI: Dependency of the stratospheric sulfur geoengineering on the aerosol microphysical model
12:00 – 13:00
Lunch
Session 3
13:00 – 13:15
Prof. Heikki Järvinen, University of Helsinki: Deterministic, stochastic, and chaotic systems
13:15 – 13:30
Dr. Aki Tsuruta, FMI, Global Carbon Project - multimodel intercomparison on global and regional GHG budgets
13:30 – 13:50
Breakout group discussion
13:50 – 14:00
Presenting group discussion results
14:05 – 14:15
Closing words for the seminar
14:15 – 14:30
Coffee break
Session 4 - The panel discussion about Climate models supporting policy makers will be in Finnish.
14:30 – 14:40
Tervetuliassanat ja johdanto paneelikeskusteluun: Ilmastomallit yhteiskunnallisen päätöksenteon tukena
14:40 – 15:30
Paneelikeskustelu
15:30 – 15:45
Yleisökysymyksiä ja yhteenveto
Dinner
18:00 – 19:30 Virtual Dinner on WonderMe
Presentation Abstracts
Dr. Laura Suarez-Gutierrez, The Max Planck Institute for Meteorology, Germany: The Max Planck institute grand ensemble: Enabling the exploration of climate system variability
The Max Planck Institute Grand Ensemble (MPI‐GE) is the largest ensemble of a single comprehensive, fully-coupled climate model currently available, with 100 members for the historical simulations and four forcing scenarios. The large ensemble size and forcing scenario diversity provide the most precise sampling of the transient internal variability in the climate system under a wide variety of global warming conditions. This makes MPI-GE a powerful tool to quantify the effect of forced changes in the climate system and separate them from internal variability, as well as to evaluate changes in the characteristics and drivers of extreme events in a warming climate. In this talk I will demonstrate how MPI-GE compares to other large ensemble experiments, based on a novel framework for evaluating how climate models capture the internal variability and forced response without the need to separate both signals in observations. Then, I will provide some examples of how we can exploit the strengths of MPI-GE, with a focus on extreme heat projections.
Dr. Juha Lento, CSC: Clean code
Current scientific process itself does nothing to encourage us to produce clean code. Only the publications are reviewed. Utterly incomprehensible or highly inefficient code will pass unnoticed through the review process. There is no reward for cleaning up the existing code. Only feature additions are appreciated, which usually are implemented as hacks, further complicating and obfuscating the code. First version of the code (a prototype, actually), that does not fail and produces plausible results, is usually also the final one. Our own professional pride or the pity for the next generation students that will inherit the code is the only motivation to ever consider the quality of the code. I see only two ways to improve the quality of the scientific codes:
Make your code truly public by publishing it in an open source repositories such as GitHub. This will almost force you to show your best effort.
Collaborate. Nobody will collaborate with you if your code is incomprehensible or inefficient. You may even get contributions beyond your own skills.
I will explain two popular programming mantras, and discuss what they mean in practice,
Keep it simple (KIS)
Don’t repeat yourself (DRY)
As an example I will show a small shell script, and what it took to make it “clean.”
MSc. Kalle Nordling, Finnish Meteorological Institute: Using CMIP6 data from Google Cloud
In this presentation, I will show two hands-on examples of how to use CMIP6 data from Google Cloud using Python. Python provides nice and easy to use api to search and browse existing CMIP6 data, without downloading the whole dataset. Example user cases here are model-to-model spread in warming trend and how daily temperatures change in different SSP scenarios, using small initial condition ensambles.
Dr. Letizia Tedesco, Finnish Environment Institute: Arctic sea ice as represented by an ensemble of CMIP5/6: findings and caveats
In this talk I will present results from an ensemble of CMIP5 and CMIP6 Arctic sea ice daily data, both from historical periods and projections scenarios. I will present new findings and caveats, and discuss progress made as CMIP moves towards new phases.
Dr. Anton Laakso, Finnish Meteorological Institute: Dependency of the stratospheric sulfur geoengineering on the aerosol microphysical model
Stratospheric sulfur injections have been proposed to complement mitigation effort to avoid catastrophic global warming. I have simulated different injection strategies with the modal and the sectional aerosol microphysical models. These results show that estimations about cooling efficiency of the stratospheric sulfur injections are strongly dependent on the model used to simulate geoengineering.
Prof. Heikki Järvinen, University of Helsinki: Deterministic, stochastic, and chaotic systems
I will present the basic concepts and examples covering deterministic, stochastic, and chaotic systems we typically encounter in weather and climate modelling problems. No new results are included, and this contribution is intended as a short pedagogical overview.
Dr. Aki Tsuruta, FMI, Global Carbon Project - multimodel intercomparison on global and regional GHG budgets
Global Carbon Project (GCP) works with the international science community to establish a common and mutually agreed knowledge of the global carbon cycle to support policy debate and actions to slow down and ultimately stop the increase of greenhouse gases in the atmosphere. Finnish scientists contributes to GCP multimodel intercomparison studies of CO2, CH4 and N2O by providing model estimates of global CH4 budgets and atmospheric concentration observations multiple tracers.