A Fast and Efficient Version of the TwO-Moment Aerosol Sectional (TOMAS) Global Aerosol Microphysics Model

Figures & data

TABLE 1 Configurations of size resolution in TOMAS-30, TOMAS-15, and TOMAS-12

TOMAS model	Aitken mode (10 nm < Dp < 100 nm)	Accumulation mode (100 nm < Dp < 1 μm)	Coarse mode (1 μm < Dp < 10 μm)
TOMAS-30	10 bins	10 bins	10 bins
TOMAS-15	5 bins	5 bins	5 bins
TOMAS-12	5 bins	5 bins	2 bins

FIG. 1 Evolution of the number size distribution predicted by the TOMAS-30, TOMAS-15, and TOMAS-12 box models, compared against the analytical solutions for (a) condensation and (b) coagulation after 1 h. To facilitate a direct evaluation of the newer algorithms, the analytical and TOMAS-30 results have been gridded to the coarser resolution of TOMAS-15. Details of the initial size distributions and other input parameters are available in Section 4. Note that the results from TOMAS-15 and TOMAS-12 are identical and thus the lines representing 2 models are overlapped. (Color figure available online.)

FIG. 2 Number size distributions predicted by the TOMAS-30 and TOMAS-15 box models using a size-dependent coagulation kernel after 1-h coagulation. Note that initial number size distribution is assumed to be lognormal with a geometric mean diameter of 30 nm and geometric standard deviations varied from 1.2 to 2.0 (σ = 1.2 for a, σ = 1.5 for b, and σ = 2.0 for c). The initial total number concentration is 10⁵ cm⁻³. (Color figure available online.)

FIG. 3 Global distribution of emission error of CN10, CN70, and CN100. Emission error is defined as errors resulting from decreasing the size resolution in the emission process and is estimated by calculating the ratio of CAD-30 to CAD-30–30EMIS. Note that the value on the top right of each figure is a global average of the errors displayed in the map. (Color figure available online.)

TABLE 2 Description of process-specific errors and simulations used to estimate the error

Error type	Error description	Simulations used
Cloud processing	Errors resulting from lower size resolution used to determine CCN activation fraction in wet deposition and in-cloud sulfur oxidation	(KOH-15/KOH-30)*
Offline CCN	Errors resulting from lower size resolution used to determine CCN(0.2%) based on model output (e.g., monthly aerosol field). This is offline process in our model	CAD-30 output merged into 15-bin/CAD-30
Emission	Errors resulting from lower size resolution used to represent emission size distribution	CAD-30/CAD-30–30EMS
Microphysics	Errors resulting from lower size resolution used to calculate condensation and coagulation rates	CAD-15/CAD-30
Condensation	Same as microphysics but no change in coagulation	CAD-15–30COA/CAD-30
Coagulation	Same as microphysics but no change in condensation	CAD-15/CAD-15–30COA
*Note that cloud-processing error is obtained by making a ratio of KOH-15 to KOH-30 and subtracting microphysics error.

TABLE 3 Descriptions of simulations used in this study

Simulation name	Condensation	Coagulation	Emission	Cloud processing/Wet deposition
KOH-30	30 bins	30 bins	15 bins	KOH
KOH-15	15 bins	15 bins	15 bins	KOH
CAD-30	30 bins	30 bins	15 bins	CAD
CAD-15	15 bins	15 bins	15 bins	CAD
CAD-12	12 bins	12 bins	12 bins	CAD
CAD-30–30EMS	30 bins	30 bins	30 bins	CAD
CAD-15–30COA	15 bins	30 bins	15 bins	CAD
Note. KOH and CAD refer to Köhler theory and constant activation diameter, respectively.

FIG. 4 Global distribution of number concentration (CN10) and mass concentrations for sulfate, sea-salt, and organic matter (shown in the 1st column) and their microphysics errors (shown in 2nd and 3rd columns) in the lowermost vertical layer. Microphysics error is defined here as errors resulting from lower size resolution used to calculate coagulation and condensation rates and is estimated by making a ratio of CAD-15 to CAD-30 (and CAD-12 to CAD-30). Note that the upper limit on the color bar is different from those in other figures because the maximum error in this figure (Figure 4) exceeds that limit and the value on the top right of each figure in 2nd and 3rd column is a global average of the errors displayed in the map. (Color figure available online.)

FIG. 5 Global distribution of microphysics error of surface layer (a) CN70, and (c) CN100, and cloud-processing error of (b) CN70, and (d) CN100, respectively. Note that the cloud-processing error is defined here as errors resulting from lower size resolution used in wet deposition and cloud processing and is estimated by making a ratio of KOH-15 to KOH-30 and then by subtracting microphysics error. Note that the lower limit on the color bar is different from those in other figures because the minimum error in this figure (Figure 5) exceeds that limit and the value on the top right of each figure is a global average of the errors displayed in the map. (Color figure available online.)

FIG. 6 Global distribution of coagulation and condensation errors of surface layer CN70, (a) to (b), and surface layer CN100, (c) to (d), respectively. Coagulation error is calculated by making a ratio of CAD-15 to CAD-15–30COAG, and the condensation error is obtained using CAD-15–30COAG and CAD-30. Note that the value on the top right of each figure is a global average of the errors displayed in the map. (Color figure available online.)

FIG. 7 Global distribution of offline CCN error of CCN(0.2%) in the lowermost vertical layer. Offline CCN error is defined here as errors resulting from decreasing size resolution used to calculate the CCN(0.2%) concentration and is based on aerosol fields from CAD-30 (see Section 5.1 for more details). Note that the value on the top right of each figure is a global average of the errors displayed in the map. (Color figure available online.)