Improving prediction of trans-boundary biomass burning plume
dispersion: from northern peninsular Southeast Asia to downwind
western north Pacific Ocean

Abstract. The boreal spring biomass burning (BB) in the northern peninsular Southeast Asia (nPSEA) are lifted into the subtropical jet stream, get transported and deposited across nPSEA, South China, Taiwan, and even the western North Pacific Ocean. This paper as part of the 7-Southeast Asian Studies (7-SEAS) project effort attempts to improve the prediction capability of the chemical transport model (WRF-CMAQ) over a vast region including the mountainous near-source burning sites at nPSEA to its downwind region. Several sensitivity analyses of plume rise are compared in the paper and it discovers that the initial vertical allocation profile of BB plume and plume rise module (PLMRIM) are the main reasons causing the inaccuracies of the WRF-CMAQ simulations. The smoldering emission from the Western Regional Air Partnership (WRAP) empirical algorithm included has improve the accuracies of PM10, O3 and CO at the source. The best performance at the downwind sites is achieved with the inline PLMRIM that accounts for the atmospheric stratification at the mountainous source region with the high-resolution FINN burning emission dataset. The calibrated model greatly improves not only the BB emission prediction over near-source and receptor ground-based measurement sites but also the aerosol vertical distribution (MPLNET, CALIPSO) and column aerosol optical depth (MODIS AOD) of the BB aerosol along the transport route. Three distinct transport mechanisms from nPSEA to the western North Pacific are then identified while a particular mechanism which involves Asian cold surge is able to mix the BB smoke plumes into the boundary layer and affects the ground surface over the western Taiwan.


The Micro-Pulse Lidar Network (MPLNET) is a federated network managed by NASA to measure the aerosol vertical structure (Welton et al., 2000). In line with the 2014 7-SEAS spring campaign conducted in nPSEA, the gridded extinction, diagnosed from the planetary boundary layer height and vertical aerosol extinction coefficient data collected is used to verify the performance of the model output (Wang et al., 2015a). The top-down lidar system, the Cloud-Aerosol Lidar with 100 Orthogonal Polarization (CALIOP) on the Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) satellite is used to study the transport pattern over larger spatial coverage to complement the single point cross-extinction profile provided by the MPLNET system. The diagnosed vertical feature mask (VFM) product is used to distinguish the aerosol types with consideration of observed backscatter strength and depolarization (Winker et al., 2011).

Anthropogenic and biogenic emission inventories
The anthropogenic emissions are re-gridded for the 1 st , 2 nd and 4 th domain (d01, d02, d04 in Fig. 1) from MIX dataset 115 available at 0.25° x 0.25° for the year 2010 Zheng et al., 2018). Model of Emissions of Gases and Aerosols from Nature (MEGAN v2.10) produces the biogenic emission input (Guenther et al., 2012) using the updated 8-day averaged leaf area index (LAI) (Yuan et al., 2011) and present-day plant functional types (PFT) from the Community Land Model version 4.0 (CLM4.0) (Oleson et al., 2010). The 3 rd domain (d03) covering Taiwan uses the 2010 anthropogenic and biogenic emissions from the locally developed Taiwan national emission database (TEDSv8.1) (TEPA, 2017). Except the 120 high quality of the East Asia national emission inventories (China, Taiwan, Japan, and Korea), large uncertainties of Southeast Asia emission due to the scarce availability of region-specific emission factor are pointed out by the inventory developers (Kurokawa et al., 2013;Li et al., 2018;Ohara et al., 2007) and local modelling efforts (Dong and Fu, 2015a;Ooi et al., 2019). Such inaccuracies are likely to affect the performance of further modeling work in the area. Therefore, energy statistics based on global anthropogenic emissions dataset, Evaluating the Climate and Air Quality Impacts of Short-Lived 125 Pollutants (ECLIPSE) developed by International Energy Agency (IEA) (Klimont et al., 2017) is used in place of the MIX dataset for peninsular SEA (PSEA). The accuracy deviation between these two datasets in nPSEA is determined through the WRF-CMAQ model performance in Section 4. The detailed comparison of ECLIPSE and MIX dataset in 2010 is discussed in Appendix B.

Biomass burning emission inventory 130
The study region is composed of small fire while small area burnt but has a rather substantial amount of fuel load and BB emissions due to the high woody compositions of the tropical and temperate forest covers. The global data set, Fire INventory from NCAR (FINN v1.5) has been applied in several previous works of literature in the region (Lin et al., 2014;Pimonsree and Vongruang, 2018) and is used as the input to the BB emission inventory into the model. A particular comparison work done for 2014 biomass burning episodes has shown FINN when used with NCEP FNL boundary condition 135 gives the greatest accuracy for PM10 at the source region compared to the GFEDv4.1 fire emission dataset (Takami et al., 2020). Seeing that the temporal speciation is handled in this research work, the main difference between fire emission inventories is the total amount of emission produced (Liu et al., 2020), hence this paper will settle with regionally more robustly tested FINN dataset for the subsequent studies. FINN is a 1 km x 1 km resolution bottom-up daily emission dataset produced from the MODIS product of active fire, land-cover type, and vegetation continuous field (Wiedinmyer et al., 140 2011). Each active fire is assumed for a 1 km 2 burnt area and the emission factor is geographically and land-cover dependent. The BB emission is processed with the fire_emis preprocessor to allocate to each grid and specify to the hourlyscale for input into the WRF-CMAQ model.

Case study setup
The plume rise module (PLMRIM) derives the initial plume top and bottom, plume rise and its dispersion according to the 145 atmospheric stability and its residual buoyancy flux (Kukkonen et al., 2014). Among a wide range of PLMRIM approaches, the simplest plume rise allocation method is the direct allocation of the initial plume top and bottom through prescribed height for all fires. This is the conventional method adopted in the case study region (Chuang et al., 2016b;Pimonsree et al., 2018). They can be determined on fixed height (Wang et al., 2013a), an empirical ratio of the plume height allocation (WRAP, 2004), adjusted with the stereo-height data from space-based Multi-angle Imaging Spectroradiometer (MISR) (Jian 150 & Fu, 2014;Val Martin et al., 2012), etc. The inline plume rise algorithm couples the interaction of BB plumes dispersion with the basic weather dynamics to determine the effective plume rise height and subsequently the plume top and bottom. This inline PLMRIM is also able to resolve the fire on the sub-grid scale and feedback the plume dynamics information into the atmospheric dynamics (Gillani & Godowitch, 1999). However, the more complex the PLMRIM gets, the higher quality and quantity of input data are required to ensure its reliability. 155 https://doi.org/10.5194/acp-2020-1283 Preprint. Discussion started: 18 February 2021 c Author(s) 2021. CC BY 4.0 License.
In this work, combinations of injection height, initial vertical distribution, smoldering fraction, and offline and inline PLMRIM are tested to determine the more suitable settings for prediction of plume rise. Five case studies are set up for the evaluation of plume rise performance and their respective initial plume rise profiles are shown in Table 2. Nofire case represents the pollution condition when no BB emission is included, while the others allocate the BB emission from the 160 FINN dataset. F800 and F2000 represent the offline PLMRIM where the injection height is fixed at generally accepted 800 m and 2000 m (Wang et al., 2013a). This fixed height method controls the plume top to be consistent hence there is no hourly and daily variation of the plume top throughout the simulation period. FWrp uses the WRAP empirical equation to allocate the initial plume rise (WRAP, 2004). The plume top and bottom vary hourly with the buoyancy efficiency with higher plume height during the hotter noontime as illustrated in the initial plume profile in Figure 2 (FWrp). However, the 165 empirical ratio adopted for each burning grid is the same every day. Idef is the inline plume-in-grid system that comes with the CMAQ model (Gillani and Godowitch, 1999). Fire emission is fed into the model at each grid point with plume top and bottom calculated through interaction of plume buoyancy efficiency and atmospheric stratification. The vertical distribution of CO plume on 12 Mar 2013 is shown in Figure 2 (Idef), but the daily weather condition is expected to vary the vertical distribution. IWrp has updated Idef with the WRAP empirical specification on burnt area size (also known as fire size). In 170 this case, the plume can be distributed according to the diurnal buoyancy efficiency and near-surface smoldering fraction as specified by WRAP. With a more reasonable BB plume peak at the noontime in Figure 2 (Hsiao et al., 2016;Pani et al., 2016) which made this site representative of the BB emissions from Myanmar, on the western side of Thailand (Khamkaew et al., 2016;Wang 190 et al., 2015a). The hourly PM2.5 data from DAK station is collected during the 2013 7-SEAS spring campaign. Table 3 shows the performance of PLMRIM on daily PM10, daily PM2.5, hourly O3 and hourly CO at LABS and DAK according to the model benchmark (correlation coefficient, R; Mean Fractional Bias, MFB; Mean Fractional Error, MFE) suggested by the Taiwan EPA (Appendix C). MFB results show that the pollutants are generally over-estimated at these mountain stations.
Unlike the case in the maritime continent that worked best with the F800 method (Wang et al., 2013a), both the fixed height 195 methods (F2000, F800) do not apply well for the nPSEA region. Only slight improvement is observed for the offline module  Table C1 for detail comparison). Adjustment of anthropogenic emission with ECLIPSE data (IWrp+EC) shows clear improvement of CO especially in the stations in Taiwan but not in Thailand. The comparably insignificant emission amount of anthropogenic emission compared to the BB emission at the near-source BB sites in Thailand is attributed to the minor pollutant changes during the BB period. 205 Among all, the inline modules (IDef, IWrp, IWrp+EC) give the lowest bias and closest correlation with the measured ground station. This highlights the importance of atmospheric stability-based PLMRIM to capture the plume rise variation at the source site. The boundary layer evolution throughout the day is very much distinctive for mountain-valley compared to the flat surface where burning usually happens. As highlighted previously (Chuang et al., 2016a;Dong and Fu, 2015b), the 210 geographical lifting mechanism at the nPSEA is the main factor the BB emission can be carried into the subtropical westerlies, and hence captured by LABS. Due to the similar performance among the offline and inline settings, the best performing setup of the offline module (FWrp) and inline module (IWRF+EC) are selected to simplify the subsequent discussion.
215 Figure 3 shows the time series plots for the hourly wind field and PM 2.5 at DAK source site and hourly wind field, PM 10, CO, and O3 at LABS. The high pollution episode (marked in grey shades) fits well with the great contrast between the model fire and nofire scenarios and thus confirming that BB plumes are the main pollution source to the high pollution episodes. overestimation when concurrent high PM10 is modelled. Short-term peak values of 4-5 hours are observed in all models for PM10, CO, and O3. The systematic peaks for these pollutants are believed to be the uncertainties involving the FINN BB emission (Pimonsree et al., 2018). It is found that the performance of O3 is relatively unaffected by the PLMRIM choice. 225

Aerosol vertical distribution
As illustrated in the shaded region in Fig. 3, the major period that affected LABS is during18-28 Mar 2013. The transport time is known to be around 2-3 days (Chuang et al., 2015), but a longer time of 4-5 days is taken to account for the BB emission generation, lifting, and dispersion at the source site. Hence, the vertical profile of the extinction coefficient from the ground lidar data on MPLNET v0 L1.5a and model output at DAK station during 13-28 Mar is compared in Fig. 4. In 240 increases with height up to around 2.3 km (775 hPa) and reduces around 3.2 km (650 hPa) before hitting another peak around 3.5-4 km. The 2-layer structure is also observed over nPSEA based on multiyear datasets from AERONET study (Feb -Apr, 2007(Gautam et al., 2013 and MISR (Feb-Apr, 2001Jian and Fu, 2014). The model output in Fig. 4c,d shows that the maximum layers above the presumed cap (3.2 km) occurred most prominently during the evening to midnight, and more often in offline than inline modules. The model shows that the offline module gives a time-invariant 255 large value over the entire layers while the inline module is giving a greater approximation on the diurnal variation with the MPLNET result throughout the day. Therefore, during the daytime, the offline module has produced a higher plume height than the CMAQ inline module (Guevara et al., 2014). However, the extinction coefficient of inline output is around 3-4 hours of time lag behind that observed by the MPLNET system.

260
The extinction coefficient from MPLNET and model output data are only available for qualitative comparison due to their generically different derivations. The lidar system determines the extinction coefficient through the backscatter feedback from the release of the laser beam at 527 nm at every minute, while, the CMAQ model used the mass reconstruction method to sum up the extinction coefficient of each model aerosol species in each layer (Mebust et al., 2003). The empirical assumption for each species and the lower vertical model resolution is attributed to the uncertainties of the modelled 265 extinction coefficient that is typically higher than the value retrieved by MPLNET.    Fig. 5a) captured the smog layer at the height of 4 km above mean sea level (amsl) over the mountainous region (Fig. 5b,c). The aerosols detected are mainly made up of smoke and mixed polluted continental aerosols, which is the main burning emission source. It is known that the burning aerosols 275 from the west part of nPSEA are orographically lifted by west-to-south-westerlies to a higher altitude depending on the terrain height (Cheng et al., 2013;Wang et al., 2015b). For the swath in Fig. 5d f, the aerosol layers are detected on high levels up to 4 km during the midday. It is most certain to be transported over from the nPSEA since the aerosol layer is detected over the sea where burning does not occur. Secondly, the plume thickness is around 4 km despite the flat land surface, which is much higher than the source site which usually ranges between 0 -3 km. The aerosol layers are believed to 280 be lifted to a higher level and also mixed to the surface over the land mask in southeastern China. This region locates one of Recently, it is proven through brute-force methods that the pollution from clusters arrived at the higher altitude in Taiwan during the winter season (Chuang et al., 2019). About 12 hours later when the swath (Fig. 5g i) moves closer to 285 Taiwan, the plumes move towards north of 16 ºN but still maintain at a similar altitude that can be detected by the LABS station at 2.4km amsl (Fig. 1). The plume is also found to continue gain in moisture content along the path.  A detailed comparison of vertical distribution for all sensitivity tests is given in Appendix D. but here we continue to discuss 295 FWrp and IWrp+EC cases. In general, the offline FWrp produces a much higher concentration of high PM10 aerosol layers compared to the inline IWrp+EC. Figure 5 shows the model PM10 result for FWrp (range: 0-300 µg m -3 ) and IWrp+EC (range: 0-120 µg m -3 ) for the corresponding period of CALIPSO swath in Fig. 5. Comparison of Fig. 6a-d shows that the FWrp produces higher plumes and IWrp+EC produces lower plumes since the former produces the initial plume profile on 19 Mar that is consistently high and less dependent on the atmospheric stability induced by mountain flow ( Figure D1). 300 Further from the source site (Fig. 6e,f), both runs predict a much lower aerosol layer around 2 km, compared to the 4 km height captured by the CALIOP sensor. The under-representation of both systems along the transport path above sea might be due to the moisture detrainment and entrainment process that is not accounted for in the current model (Paugam et al., 2016;Sofiev et al., 2012).

305
With a concentration difference of more than 2 times between FWrp (up to 300 µg m -3 ) and IWrp+EC (up to 120 µg m -3 ) , a more accurate value is captured at LABS by the IWrp+EC as shown in Table 3. Regardless of the PLMRIM used, the top height of the plume is confined by an overhead upper-layer wind system. The system has created a strong shear and suppressed the lifting pertaining to the burning convective heat. This explains the invariant of plume height when different settings are used. 310 The cross-sectional profile in Fig. 6 shows that the amount of emission produced by the offline method is substantially larger than the amount produced by the inline method. Therefore, the total columnar AOD data provided by 1º x 1º MODIS Terra 315    Figure 7 shows that the total column AOD produced by the inline module gives a closer approximation to the MODIS. FWrp greatly overestimates the aerosol produced by the BB emissions, while the inline module gives a closer agreement on northern Thailand and southern Vietnam.

Reliability of inline PLMRIM
The variation of model performance has intrigued the compatibility of emission inventory with the PLMRIM performance.
The FINN dataset provides high-resolution data for each fire (1 km 2 ) and would be more representative in the inline 325 calculation that is proceeded with the plume-in-grid concept. Therefore, if the offline method is adopted (FWrp), the highresolution emission dataset FINN in the nPSEA region tends to over-predict by 4-fold (Fig. 3a). Previous literature has to make an adjustment to the fire inventory to bring down the FINN emission amount that was overestimated by up to 2-3 times of PM2.5 and PM10 at the source region (Pimonsree et al., 2018), and FLAMBE overestimates up to 3 times for CO and PM10 at the LABS site (Chuang et al., 2015;Fu et al., 2012). In this paper, the model discovers that the direct application of the 330 FINN dataset is able to work well with the inline module (IWrp+EC). BB emission is mainly caused by small fires and dry conditions over the period in the region (Giglio et al., 2013;Reid et al., 2013), this also explains why the inline module worked well to represent the BB condition.  The inaccuracy of the offline module is likely to be caused by the role of the complex terrain in uplifting the smoke plume and the nature of the fuel loadings. The connecting slopes (0.2-1.8 km as seen in Fig. 1c) causes the complication to boundary layer physics that governs the dynamics to transport the plumes formed in the valley pockets. Due to the unique topographic structure in nPSEA, the lifting and breaking away of burning emission plumes from burning area occurs during 340 the evening-to-night period. Therefore, mountain meteorology played an important role in the distribution of higher-level plumes. Moreover, the ability of PLMRIM to capture the boundary layer physics becomes essential in the mountainous region. Through the inline module with the WRAP initial plume profile (IWrp+EC), the natural buoyancy of fire together with the convective interaction of the atmosphere can correctly distribute the BB emission. The spatial distribution of PM 10 over burning regions in nPSEA is shown, with comparison made for scenarios nofire (Fig. 8a), offline (Fig. 8b) and inline 345 (Fig. 8c). Comparison of the figures shows that each sub-grid scale fire hotspots more realistically represents the actual high concentration of emission emitted at the source (Fig. 8c) compared to the grid-following averaged out effect in the offline method (Fig. 8b). Nevertheless, the current setting does not include the two-way aerosol-radiation and aerosol-radiationcloud feedback. This will be further studied in the future work looking at its importance in the cloud-laden SEA region (Tsay et al., 2016), as seen in the missing data due to the cloud cover in Fig. 6d. 350

Transport of biomass burning aerosol to Taiwan
The below discussion is performed using the model output of IWrp+EC and focuses on the high pollution episodes observed at LABS during 13-28 Mar 2013 as seen in the grey shaded area of Figure 3. In the source region of nPSEA, the complex land terrain has played a substantial role in the BB plume lifting. Figure 9 shows the evolution of the PM10 355 concentration on 13 Mar 2013 at DAK but over the nPSEA through the cross-sectional profile (Fig. 1c). During the day (a) Nofire (b) FWrp (c) IWrp+EC when the fires are active, BB emission is released from the surface (Fig. 9a, b). Along with the rising of planetary boundary layer height (PBLH), the BB aerosol mixes into the entire boundary layer. The residue layer starts to form during the transitional period between the day and night around 17:00 LST (Fig. 9c) when the ground surface cools down. When the atmosphere becomes stable into the night, the aerosol layer remains as the residue layer and does not move down with the 360 boundary layer (Fig. 9d). The plume starts to be advected by the shear of the upper layer flow at night on the downwind leeside of the hills. It is because the boundary layer height tends to rise higher due to turbulence. The descent of the boundary layer also confines the aerosol and causes a high concentration near the surface. The detachment of the aerosol layer therefore explains the two-layer plume feature from evening into the night in Fig. 4b,c. The dispersion of emission from the pockets is subjected to at least three systems, (i) strong westerlies from Myanmar flowing over the top of valley 365 pockets that confined the emission (terrain structure shown in grey in Fig. 9), (ii) diurnal mountain-valley breeze might trap or disperse the emission, (iii) local heating caused by the solar cycle affects the plume rise and disperse the emission.
Therefore, the amount of burning emission lifted is greatly coherent with the populated hills along the transport path.
While BB contributes 43±31% to PM10, 41±32% to PM2.5, 23±19% to O3, and 39.1±23.0% to CO at LABS for the entire 375 month of Mar 2013. The transport pathway of BB from nPSEA to LABS coincides with the anthropogenic emissions from the nPSEA as well as the southeast China, BB aerosols from such emission region are also captured in the model. Therefore, the actual amount might indicate a slightly lower contribution by BB aerosol than the derived contribution. There are several mechanisms identified in Mar 2013 to bring BB smoke to Taiwan.

Westerlies to carry BB emission to LABS 380
In this case, the BB aerosol lifted is further carried by strong westerlies on the upper layer, around height between 2-4 km towards LABS. This usually occurred during the night when the atmospheric boundary layer is low and stable as shown in Fig. 10. This is the commonly known mechanism that carries the BB plumes to higher ground in Taiwan. This condition occurred on 19-20, 24-25, 27-28 Mar 2013. This is the commonly known scenario that is well studied due to the availability of measurement collected at LABS. 385

Mixing of BB emission with local pollution on surface
The land surface is heated up and the boundary layer during the day grows as high as 1.5 -2 km on western Taiwan, around 390 1 km on the windward of the central mountain range, and up to 4 km amsl at LABS. When the BB plumes overpass are as low as the BLH, then the BB aerosol is brought into the boundary layer and mixed to the ground as shown in Fig. 11. The interaction of BB with local pollutants depends on the loading of local pollutants present. The latter is subjected to the local weather system and the occasional Asian continental cold surge that might clean the accumulated pollutants. Such cases usually occur during the morning to noontime when the land surface heats up and PBLH develops. This condition occurred 395 on 18, 19, 20, 21, 28 Mar 2013. This is the main mechanism that affects the western Taiwan. It was pointed out that cold surge might be responsible for the downdraft of the BB smoke plumes to the surface (e.g. Lin et al., 2017).

Mixing of BB emission with local pollution above surface
Along with the sea-land heat difference, the sea breeze and mountain breeze are formed and enhance the uphill movement of local pollution in western Taiwan. In such a case, the local pollution is brought up to a high elevation to interact with the BB smoke plumes as shown in Fig. 12. It also occurred that the local pollutants brought uphill detaches from the planetary 405 boundary layer when the surface cools down quickly. This residue layer of pollutants is then mixed into the BB layers and carried towards the east. Such cases usually occur during midday when the local pollution plumes have moved up to the hill. This condition occurred on 17, 23, 25 Mar. The detection of BB intrusion into surface sites in southwestern Taiwan is not a rare occasion Tsai et al., 2012). A larger amount of fine nanoparticles from local sources is measured at LABS especially during the morning even not during the spring season . Therefore, it is possible that 410 mixing does occur when the local pollutants are transported up the hill through the valley breeze.

415
Among the three mechanisms, the BB aerosols have the most direct influence on the surface site in western Taiwan, which is coherent to the reduction of surface O3, NOx, and SO4 2aerosols in 2006 (Dong et al., 2018). However, all these three mechanisms are prone to alter the radiative forcing over western Taiwan. The future incorporations of the aerosol radiative forcing effect through one-way and two-way meteorology-chemistry process of moisture detrainment and entrainment are necessary to understand the role of BB aerosol on the weather extremes in downwind regions. The cloud-aerosol interaction 420 is particularly crucial to the study of the impact of BB aerosols on cloud-laden regions between nPSEA and Taiwan (Hsu et al., 2003;Tsay et al., 2016). The allocation fraction will need to improve looking at the importance of small fire smoldering in SEA (Akingunola et al., 2018;Zhou et al., 2018). The model comparison shows that regardless of the injection height, the main deficiency of the fixed height offline algorithm 435 originates from its invariant vertical-layer allocation of BB concentration throughout the day. In the complex terrain over the nPSEA region which is continuous and varies between 0.2 km to 1.8 km, mountain meteorology played an important role in the distribution of higher-level plumes. The two-layer structure of the BB plumes observed in the MPLNET extinction coefficient profile at night is well captured by the inline PLMRIM (IWrp+EC) while the offline method (FWrp) gives a time-invariant large value over the entire layers. This highlights that the inline PLMRIM (IWrp+EC) is able to incorporate 440 the diurnal boundary layer physics of the mountain to accurately represent the vertical distribution of the BB concentration in the source and downwind region. It is then clear that the amount of emission produced by the inline reasonably captures the columnar AOD distribution over the transport route between nPSEA and downwind Taiwan when compared to the MODIS columnar product. It is discovered that the inline module with the initial distribution profile of WRAP (IWrp+EC) is able to and performs well both at the source and receptor sites compared to the offline module. 445 The model output shows that the BB plumes near nPSEA are emitted during the day within the BLH. Due to strong mountain-valley wind, the smoke plume layers tend to detach from the BLH as residue layers when the surface cools down in the evening-to-night period. This is the layer of plumes that entered the free troposphere at approximately 1-3km height and further transported over to western north Pacific and Taiwan. The plume layers clearly affect the Taiwan region via three 450 conditions: (a) overpass western Taiwan and enter mountain area (LABS), (b) mix down to western Taiwan, (c) transport of local pollutants up and mix with BB plume on LABS. The second condition involves the prevailing high-pressure system that is able to impact the most population in Taiwan and would be an interesting case to explore.
However, care should be taken to select the BB emission inventory input when switching from the offline module to the 455 inline module. The sub-grid scale allocation of the BB emission requires higher resolution of BB emission inventory such as stability-based PLMRIM and the accurate application of emission inventories to capture the plume rise variation at the source site with complex terrain. The correct representation at the nPSEA source site substantially affects the downwind BB concentration in mountain (LABS) and surface sites in Taiwan. It is also observed that the improved setting is able to 460 represent the source site's vertical profile well, however, the height of the plume is reduced following the transport and evolution of the plume approaching Taiwan. This might be caused by the missing algorithm of the indirect and direct effect between aerosols and the high cloud cover region along the transport path. It leads to future exploration and incorporation of the effect of cloud-aerosol interaction over the cloud-laden region. The boundary condition data in WRF model uses the reanalysis weather data. These data are assimilated with measurement data, they are available in coarse resolution (1° x 1°). The work has hence included the observation nudging settings to 475 improve its prediction of local area. The data used for nudging are given in Section 2. The assimilation with the default setting does not improve the prediction hourly T2 and WS, hence the subsequent effort is to adjust the area of influence of each the measuring stations. The radii of influence (RIN) for both d03 and d04 are updated to 100 km based on the average distance between the observation stations (d03: 125 km, d04: 153 km) and minimum distance between 2 stations (d03: 64 km, d04: 36 km). Although the wind direction is greatly improved with the modification of RIN, the positive bias of T2 and 480 negative bias of WS is still apparent, especially for the LABS station. Given that the 3 rd domain is of 5 km x 5 km resolution, the height of Mt. Lulin might be averaged out by the lower terrain surrounding it and the model height of Mt. Lulin is lower (2216 m, layer = 1) than its original height (2862 m). Comparison has found that model layer 4 from surface is most representative of the height of Mt Lulin (2492 m; 757 hPa). Hence with the extraction of new location of Mt Lulin, the prediction of T2 and WS are improved significantly as tabulated in $Table_VERmet. The wind profile over LABS, one of 485 the decisive weather factors of transport, has complied well with the observation data as seen in Figure 2. The passing rate of surface cwb stations for hourly T2, WS and WD are also well above the model benchmark (60%).

Appendix B. Comparison of ECLIPSE and MIX anthropogenic emission
The anthropogenic dataset, ECLIPSE and MIX for year 2010 is compared in Figure B1 for peninsular SEA and in Figure B2 for the entire Asia. Figure B1 shows that ECLIPSE generated lower amount of CO and VOC and higher amount of particulate matters and NOx over peninsular SEA compared to the MIX dataset. The ECLIPSE data give a higher total NH3, BC, PM2.5, NOx, PM10 by 192%, 51%, 38%, 29%, 24% respectively, while lower total VOC, CO, OC, SO2 by 40%, 23%, 495 22%, 20% respectively. Largest biases are observed in developing SEA countries as seen in Figure B2, such as Laos, Burma, Philippines and Timor-Leste where local data are not easily available. However, the emissions for China and Taiwan are kept unchanged due to the high confidence and quality of respective national emission inventories .

Appendix D. Detailed comparison of vertical distribution 515
For offline methods, higher plume rise height and concentration vary positively with the initial allocated height (Table 2), with increasing order of F800, F2000 to FWrp. Inline method is generally lower in amount and the near surface emission has increased with IWrp compared to IDef ( Figure S2).

Data availability
All the data sets presented in this study are available upon request from the corresponding author.