BIOMASS BURNING IN THE NASA GISS MODELE: UNCERTAINTY AND INTERACTIONS BETWEEN EMISSIONS, TRANSPORT AND CHEMISTRY

ACCURATE MODEL SIMULATION OF THE EFFECTS OF BIOMASS BURNING ON THE CHEMICAL AND CLIMATE SYSTEM DEPENDS ON THE EMISSIONS ESTIMATES IN THE REPRESENTATION OF THE UNDERLYING CORE CLIMATE PROCESSES AND IN THE REPRESENTATION OF CHEMICAL PROCESSES. CHANGES IN A MODEL S CLOUD SCHEME FOR EXAMPLE WILL AFFECT THE VERTICAL TRANSPORT AND DISTRIBUTION OF EMISSIONS THE CHEMICAL PROCESSING OF THOSE EMISSIONS VIA CHANGES IN MOISTURE AND CLOUD COVER AND THE WET REMOVAL OF SOLUBLE SPECIES. WHEN A MODEL USES A PROGNOSTIC FIRE SIMULATOR THE FLAMMABILITY AND THEREFORE THE PREDICTED BIOMASS BURNING EMISSIONS WILL ALSO DEPEND STRONGLY ON THE MODEL S CLOUD PROCESSES. UNDERSTANDING AND REDUCING MODEL-OBSERVATION DISCREPANCIES IN ATMOSPHERIC COMPOSITION HAS FOCUSED LARGELY ON BIASES IN PRESCRIBED BIOMASS BURNING EMISSIONS AND THE REPRESENTATION OF CHEMICAL PROCESSES. LITTLE EFFORT HAS BEEN SPENT UNDERSTANDING THE CONTRIBUTION OF CORE CLIMATE PROCESS UNCERTAINTY TO DISCREPANCIES IN ATMOSPHERIC COMPOSITION. TO THE BEST OF OUR KNOWLEDGE THIS HAS NOT BEEN EXAMINED SYSTEMATICALLY ALONGSIDE UNCERTAINTIES IN EMISSIONS AND CHEMICAL REPRESENTATIONS. BASED ON OUR RECENT WORK CORE MODEL PHYSICS CHANGES CAN INDEED HAVE EFFECTS COMPARABLE TO EMISSIONS OR CHEMISTRY ON SIMULATED ATMOSPHERIC COMPOSITION. BECAUSE OF THIS WE ARGUE THAT CHANGES TO EMISSIONS OR CHEMISTRY SCHEMES THAT PRODUCE AN APPARENT IMPROVEMENT IN PERFORMANCE COULD ACTUALLY BE COMPENSATING FOR ERRORS ELSEWHERE IN THE MODEL. CONSEQUENTLY WE ARGUE THAT MODEL-BASED ESTIMATES OF THE EFFECTS OF BIOMASS BURNING ON THE CHEMICAL AND CLIMATE SYSTEM RAINFALL PATTERNS FOR EXAMPLE - ARE MOSTLY MODEL-DEPENDENT MAKING IT DIFFICULT TO DRAW ROBUST CONCLUSIONS. THIS IS ESPECIALLY THE CASE AS MODELS BECOME MORE SOPHISTICATED WHICH IS ALWAYS THE CASE WITH MODELE. TO ADDRESS THIS PROBLEM WE PROPOSE TO CONDUCT PERTURBED PHYSICS ENSEMBLES (PPES) WITH THE NASA GISS MODELE COUPLED CHEMISTRYCLIMATE MODEL. AS PART OF PROPOSED WORK WE WILL ENHANCE THE CURRENT MODELE FIRE SIMULATOR AND IMPLEMENT A SIMPLE INJECTION HEIGHT MODEL. PPES WILL BE CONDUCTED ACROSS THE PROGNOSTIC FIRE INJECTION HEIGHT SUBGRID ATMOSPHERIC PHYSICS AND COMPOSITION COMPONENTS OF MODELE. PPE MEMBERS WILL BE EVALUATED JOINTLY AGAINST A REPRESENTATIVE SET OF COMPOSITION AND CLIMATE OBSERVATIONS. A REDUCED SET OF PERMISSIBLE PPE MEMBERS WILL BE USED TO EXPLORE OUTSTANDING QUESTIONS RELATED TO THE SOURCES OF CARBONACEOUS AEROSOLS IN BIOMASS BURNING REGIONS TO MECHANISTICALLY UNDERSTAND THE TRANSPORT AND CHEMICAL PATHWAYS THROUGH WHICH BIOMASS BURNING AFFECTS ATMOSPHERIC COMPOSITION AND REGIONAL CLIMATE AND TO PRODUCE A RANGE OF FUTURE BIOMASS BURNING IMPACTS SCENARIOS MORE COMPREHENSIVELY ACCOUNTING FOR MODEL UNCERTAINTY.