New developments in the General NOAA Operational Modeling ......New developments in the General NOAA...

New developments in the General NOAA Operational Modeling Environment (GNOME)

Christopher H. Barker NOAA Office of Response and Restoration Emergency Response Division http://response.restoration.noaa.gov Chris.Barker@noaa.gov

Outline  • ORR  /  ERD  background  • GNOME  /  ADIOS  History  •  ComputaAonal  Structure  /  Features  • Oil  Database  •  ScripAng  /  Extending  • Web  Interface  

NOAA Emergency Response Division •  US  National  Contingency  Plan:  

NOAA  to  provide  Scientific  Support  Support  for  Hazardous  Materials  spill  response  in  Coastal  Waters  

•  “All  Hazards”  response  support  for  over  35  years  

•  Provide  support  for  100-­‐200  responses/year  •  In-­‐house,  interdisciplinary  team  of  spill  scientists  

•  A  24/7  National  network  of  expert  support  

ERD  Modeling  •  Model  development  in-­‐house  •  Model  developers  and  operaAonal  modelers  are  

the  same  team  •  We  understand  the  strengths  an  limitaAons  •  We  can  add  features  /  fix  bugs  on  the  fly  •  Tools  are  developed  specifically  to  support  

response  

Oil  Spill  Response  Model  Requirements  

•  Quick  to  iniAalize:  Answers  within  hours  •  Easy  to  Calibrate:  What  if  the  results  do  not  match  the  field  observaAons?  

• Wide  range  of  scales  •  Can  ingest  whatever  is  available:  

– HF  Radar,  circulaAon  models,  etc.  •  And  not  just  oil:  marine  debris,  whale  carcasses,  larvae,  bodies…  

GNOME  1:  • C++  • Desktop  GUI  

(Windows  and  Mac  Carbon)  • Code  is  tightly  integrated  • Primarily  Transport  • Weathering  Completely  Independent  (ADIOS)    

•  Limited  Batch  Mode  

GNOME  2  Goals  • Web  interface  •  Full  scripting  interface  •  Integrated  weathering  •  Easier  to  add  new  features  

–   Plug  in  your  own  Movers,  Weatherers,  Maps,  Element  Types  

•  Easier  to  test/maintain/improve  • Open  Source  development  model  

GNOME  Key  Features:  • Particle  Tracking  (Lagrangian  Elements)  

• Linear  Superposition  of  Physical  processes  

GNOME  Key  Components  • Movers  

Anything  that  moves  an  element  • Weatherers  

Changes  the  element’s  properties  • Maps  

Interaction  with  shoreline/bottom  •  Releases  

When  and  where  (and  what)  elements  are  injected  

Wide  Range  of  Problems  •  Subsurface  Blowout  

– Fully  3-­‐d  – Weathering  of  multi-­‐component  oil  

•  Tsunami  Debris  – Surface  – Wind  drift  only  – Very  large  time/spatial  scale  

•  Larva  modeling:  – 2  or  3-­‐d  – Larval  behavior  

GNOME  Inputs  •  Currents:  

–  Pattern  scaled  by  time  series  –  Pattern  scaled  by  components  of  wind  

–  Circulation  models:  •  Rectangular  grids  •  Curvilinear  grids  •  Unstructured  grids  

•  Winds:  –  Time  series  at  a  point  –  Gridded  winds  from  met  models  

Spill  Sources  •  Oil  Types:  

•  Non-­‐weathering  •  Simple  oils  •  Full  ADIOS  database  

•  Plume  model  •  Point  or  line  •  Instantaneous  or  continuous  •  Can  superimpose  multiple  spills  

Fate  /  Weathering  •  Pseudo-­‐component  model  

– Distillation  cuts  (SARA)  • Saturates  • Aromatics  • Asphaltines  • Resins  

•  ~1000  oils  in  database  – Estimations  for  missing  properties  – “Generic”  Oils  

Weathering  Algorithms  • Spreading  • Evaporation  • Natural  Dispersion  • Sedimentation  • Emulsion  formation  • Biodegradation  • Dissolution  

Weathering  Algorithms  • Spreading  • Evaporation  • Natural  Dispersion  • Sedimentation  • Emulsion  formation  • Biodegradation  • Dissolution  

New!  

Weathering  Algorithms  • Spreading  • Evaporation  • Natural  Dispersion  • Sedimentation  • Emulsion  formation  • Biodegradation  • Dissolution  

And  Improved!  

Response  OpAons  • Skimming  • Dispersant  Application  • Burning  

All  really  simple  user-­‐specified  effectiveness  

GNOME    ScripAng  Interface:  

•  Fully  Python  Scripted  •  Batch  processing  •  Add  your  own  front  end  •  Add  your  own  movers,  etc  in  

Python  (or  any  code  called  from  Python)  

GNOME  Web  Interface:  Running  on  NOAA  server  

•  Simple  use  (location  files)  •  Intuition  building  •  Educational  •  Download/save  your  setup.  

Run  your  own  server:                (could  be  your  laptop)  

•  Custom  locations,  etc.  

GNOME  Transport  Mode  

Oil  Budget  

Oil  Database  

GOODS  

ORR web site: http://response.restoration.noaa.gov

WebGNOME: http://gnome.orr.noaa.gov

The Source Code: https://github.com/NOAA-ORR-ERD !Chris Barker NOAA Office of Response and Restoration Emergency Response Division Chris.Barker@noaa.gov