BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20230831T095745Z
LOCATION:Sertig
DTSTART;TZID=Europe/Stockholm:20230626T120000
DTEND;TZID=Europe/Stockholm:20230626T123000
UID:submissions.pasc-conference.org_PASC23_sess177_pap134@linklings.com
SUMMARY:Approximation and Optimization of Global Environmental Simulations
  with Neural Networks
DESCRIPTION:Paper\n\nElnaz Azmi, Jörg Meyer, and Marcus Strobl (Karlsruhe 
 Institute of Technology); Michael Weimer (Massachusetts Institute of Techn
 ology); and Achim Streit (Karlsruhe Institute of Technology)\n\nSolving a 
 system of hundreds of chemical differential equations in environmental sim
 ulations has a major computational complexity, and thereby requires high p
 erformance computing resources, which is a challenge as the spatio-tempora
 l resolution increases. Machine learning methods and specially deep learni
 ng can offer an approximation of simulations with some factor of speed-up 
 while using less compute resources. In this work, we introduce a neural ne
 twork based approach (ICONET) to forecast trace gas concentrations without
  executing the traditional compute-intensive atmospheric simulations. ICON
 ET is equipped with a multifeature Long Short Term Memory (LSTM) model to 
 forecast atmospheric chemicals iteratively in time. We generated the train
 ing and test dataset, our target dataset for ICONET, by execution of an at
 mospheric chemistry simulation in ICON-ART. Applying the ICONET trained mo
 del to forecast a test dataset results in a good fit of the forecast value
 s to our target dataset. We discussed appropriate metrics to evaluate the 
 quality of models and presented the quality of the ICONET forecasts with R
 MSE and KGE metrics. The variety in the nature of trace gases limits the m
 odel's learning and forecast skills according to the respective trace gas.
  In addition to the quality of the ICONET forecasts, we described the comp
 utational efficiency of ICONET as its run time speed-up in comparison to t
 he run time of the ICON-ART simulation. The ICONET forecast showed a speed
 -up factor of 3.1 over the run time of the atmospheric chemistry simulatio
 n of ICON-ART, which is a significant achievement, especially when conside
 ring the importance of ensemble simulation.\n\nDomain: Climate, Weather an
 d Earth Sciences\n\nSession Chair: William Sawyer (ETH Zurich / CSCS)
END:VEVENT
END:VCALENDAR