Sensitivity of biomass burning emissions estimates to land surface information

. Emissions from biomass burning (BB) are a key source of atmospheric tracer gases that affect the atmospheric carbon cycle. We estimated four types :::::::: developed :::: four ::: sets : of global BB emissions ::::::: estimates ::::::: (named :::::::: GlcGlob, :::::::: GlcGeoc, ::::::::: McdGlob, ::: and ::::::::: McdGeoc) : using a bottom-up approach and by combining the remote sensing products related to ﬁre distribution with two aboveground biomass (AGB) and two land cover classiﬁcation (LCC) distributions. The sensitivity of the estimates of BB emissions to the AGB and LCC data was evaluated using the carbon monoxide (CO) emissions associated with each BB 5 estimate. We found a substantial spatial difference in CO emissions for both ::::: Using : the AGB andLCC data, which resulted in :: /or ::::: LCC :::: data ::: led :: to ::::::::::: substantially ::::::: different :::::: spatial :::::::: estimates :: of :::: CO ::::::::: emissions, :::: with : a large (factor of approximately three) spread of estimates for the mean annual CO emissions : ; :::::::: 526 ± 53 , :::::::: 219 ± 35 , :::::::: 624 ± 57 , :::: and ::::::: 293 ± 44 ::: Tg ::: CO ::::: yr − 1 , ::: for :::::::: GlcGlob, :::::::: GlcGeoc, :::::::: McdGlob, ::: and ::::::::: McdGeoc, ::::::::::: respectively, ::: and ::::::: 415 ± 47 ::: Tg ::: CO :::: yr − 1 ::: for :::: their :::::::: ensemble ::::::: average :::::::: (EsmAve). We simulated atmospheric CO variability : at ::::::::::::: approximately :::: 2.5 ◦ :::: grid : using an atmospheric tracer transport model and the BB emissions 10 estimates and compared it with ground-based and satellite observations. At ground-based observation sites during ﬁre seasons, statistical comparisons indicated that the impact of differences in the BB emissions estimates on atmospheric CO variability ::::::::: intermittent ::: ﬁre :::::: events was poorly deﬁned in our simulations ::: due :: to ::: the ::::: coarse ::::::::: resolution, ::::: which :::::::: obscured :::::::: temporal ::: and :::::: spatial :::::::: variability :: in :::: the :::::::: simulated :::::::::: atmospheric :::: CO :::::::::::: concentration. However, when compared at the regional and global scales, the distribution of atmospheric CO concentrations in the simulations show substantial differences among the estimates of BB emissions. These results indicate that the estimates of BB emissions are highly sensitive to the AGB and LCC data. :::::::: :::: :::: ::::::::: :::::: ::::: :::::: ::::: :::::: :: of ::: BB ::::::::: :: at :::: :: at ::: the :::: globe ::::: scale: ::::: large :::::::: in ::: :::::::::: ::: ::::::::: ::::: :::::: ::: ::::::::: :::::::: ::::: ::::::::::: comparisons :::::: : a :::::::: difﬁculty ::::: faced ::: by ::: the :::::::::: simulations :: in ::::::::::: reproducing ::: the :::::::::::: ground-based :::::::::: observations ::: of ::::: higher ::: CO ::::::::::::: concentrations :::::::: generated :: by :::::: sudden :::: BB :::::::: emissions :::: from :::::::::: intermittent ::: ﬁre :::::: events.


55-year
to simulate substantial atmospheric transport. We used the NICAM-TM version described by Niwa et al. (2017) for the transport of CO.
Fossil fuel, biogenic, and biomass burning CO-emission-inventories were used as the CO emission sources at the Earth's surface. The fossil fuel CO emissions were derived from the Emissions Database for Global Atmospheric Research (EDGAR v4.3.2;Janssens-Maenhout et al., 2019) with an annual resolution. Biogenic CO emissions from vegetation were derived from a process-based model, the Vegetation Integrative SImulator for Trace gases (VISIT; Ito, 2019). The biogenic CO emissions in 125 VISIT are simulated as a part of processes associated with biogenic volatile organic compound emissions and have a monthly resolution. For CO emissions from BB, the abovementioned four scenarios :: are based on the various combinations of the LCC and AGB maps.

Modified index of agreement and standardized anomaly
We used the modified index of agreement (MIA) (Willmott et al., 1985) to compare the observed and simulated atmospheric CO concentrations, as follows: 185 where x and y are the observed and simulated CO concentrations (ppm) and x is the sample mean of x. The MIA calculates normalized value from 0.0 to 1.0 with higher values indicating better agreement between the observations and the model simulations. Correlation coefficient indicates higher value for agreement of phase variations in the variability, whereas the MIA does for both agreements of phase and amplitude gain variations in the variability.
The observational time series from the BKT and ETL sites were used to classify the 'no fire' or 'fire' months based on the 190 standardized anomaly z: where x is the observed daily CO concentration (ppb) and σ x is the corresponding sample standard deviation. In this study, fire months were empirically identified as having observed CO concentrations corresponding to z i ≥ 1.5.
comparison. The forest area in GLC2000 was 55.8 × 10 6 km 2 , 199% more than that in MCD12Q1 (28.0 × 10 6 km 2 ); the area of shrub/savanna/grass in GLC200 is 56.4 × 10 6 km 2 , 43% less than MCD12Q1 (98.6 × 10 6 km 2 ); that of crop in GLC2000 was 28.2 × 10 6 km 2 , 181% more than MCD12Q1 (15.6 × 10 6 km 2 ). At the global scale, it is noteworthy that there are large differences in the area totals of the vegetation classes between the two products; e.g., GLC2000 possesses larger forest areas, 215 whereas MCD12Q1 has more shrub/savanna/grass. Giri et al. (2005) found that the spatial distribution of vegetation in eight LCC classes shows agreement of 59.5% between the GLC2000 and MCD12Q1 products, and the discrepancies between them occur in southern Siberia, the Sahel region, southeastern Brazil, Southern Australia, and on the Tibetan plateau.
Furthermore, although it is clear that there are significant differences among the various land surface information products currently available, the quantitative evaluation of these differences remains difficult because of the limited coverage of surface observations. One approach to addressing this limitation would be the commissioning of future satellite missions carrying 455 higher-resolution onboard sensors.