Solar radiative forcing of aerosol particles near the Taklimakan desert during the Dust Aerosol Observation-Kashi campaign in Spring 2019

The Taklimakan desert is a main and continuous source of Asian dust particles causing a significant direct aerosol solar radiative forcing (ASRF). In order to improve the accuracy of the estimation of dust radiative forcing effects, the Dust Aerosol Observation-Kashi (DAO-K) campaign was carried out near the Taklimakan desert in April 2019. The objective of the campaign is to provide comprehensive parameters such as: dust optical and microphysical properties, vertical distribution 15 and surface albedo, for the calculation of ASRF. The measurements were employed in radiative transfer (RT) simulations and the estimations are improved by considering the actual measured atmospheric profiles and diurnal variations of land surface albedo in addition to reliable aerosol parameters. The RT model estimates the ASRF results in average daily mean cooling effects of -19 W m at the top of atmosphere and -36 W m at the bottom of atmosphere during the DAO-K campaign. The Weather Research and Forecasting model with Chemistry (WRF-Chem) with assimilations of the aerosol optical depth and 20 PM2.5 and PM10 concentrations measurements is prone to overestimate the radiative forcing effects of dust aerosols. The percent difference of daily mean ASRF between the two simulations are greater than 50% in heavy dust episode. Ground-based observations of downward irradiances have validated that the RT simulations are in good agreement with simultaneous observations, whereas the WRF-Chem estimations exhibit obvious discrepancy with these independent measurements. Data assimilations can partly reduce the discrepancy, but there is still room for improving the WRF-Chem simulation of dust aerosol 25 radiative forcing.

The paper begins with a brief introduction of the Dust Aerosol Observation-Kashi (DAO-K) field campaign, and an introduction of the multi-source observations and data in Sect. 2. Methods to estimate ASRF by improving the input parameters in RT simulation and by employing data assimilations in the WRF-Chem simulation are described in Sect. 3. Sect. 4 presents the instantaneous and daily mean results of ASRF simulated by RT model during the DAO-K field campaign. The influences 70 of the atmosphere and surface conditions on the results are discussed. They are also evaluated by comparing with AERONET operational products, WRF-Chem simulations, and simultaneous irradiance measurements. Summary and conclusions are given in Sect. 5.

Experimental site and instrumentation 75
The Dust Aerosol Observation-Kashi (DAO-K) field campaign with comprehensive observations of physical, chemical, and optical properties of aerosol particles, radiative properties, vertical structures of atmosphere, and land surface albedo in this region was designed to provide high quality data for aerosol radiative forcing estimates. Kashi is located in the vicinity of the Taklimakan Desert; it is surrounded by the Tianshan Mountains in the north, the Pamir Plateau in the west, and the Kunlun Mountains in the south (Fig. 1). Kashi represents a typical place affected by dust aerosols, local anthropogenic pollution, and 80 pollution transported from surrounding arid and desert areas. The DAO-K field campaign was conducted at the Kashi campus of the Aerospace Information Research Institute, Chinese Academy of Sciences (39.5˚N, 75.9˚E, 1320 m above mean sea level). The campus hosts a long-term observation station within the Sun-sky radiometer Observation NETwork (SONET, www.sonet.ac.cn) (Li et al., 2018). According to the SONET measurements during more than six years, the Kashi site is frequently affected by dust particles, where the multi-year average AOD is up to 0.56±0.18 at 500 nm, moreover, the Ångström 85 exponent (AE, 440~870 nm) and fine-mode fraction (FMF) at Kashi are the lowest among all sites in China. Every year, FMF reaches the lowest value, and the volume particle size distribution presents a pre-dominant coarse mode from March to May, due to the frequent dust invasions in spring (Li et al., 2018). The DAO-K intensive field campaign was carried out in April 2019 and lasted for nearly a month. During the campaign, several dust processes were observed by coordinated deployment of multiple in-situ and remote sensing platforms and instruments based on passive and active detection technologies. 90 https://doi.org/10.5194/acp-2020-60 Preprint. Discussion started: 17 February 2020 c Author(s) 2020. CC BY 4.0 License.  ground-based and satellite retrieved parameters applied in ASRF estimation and evaluation in this study. Four sun-sky radiometers, including a polarized sun-sky-moon radiometer CE318-TP (#1150), two unpolarized sun-sky-moon radiometers 95 CE318-T (#1098 and #1141), and a polarized sun-sky radiometer CE318-DP (#0971), were deployed to derive the volume aerosol properties of AOD, AE, SSA, and asymmetry factor (i.e., g) for model inputs (Li et al., 2014(Li et al., , 2015. All the radiometers have been calibrated rigorously before the field campaign. CE318 #1150 and #1141were calibrated at AERONET Izaña Observatory by Langley plot approach with the accuracy of AOD about 0.25%~0.5% at the visible and near-infrared bands and calibrated by integrating sphere with the uncertainty of radiance about 3%~5% (Holben et al. 1998;Li et al., 2018). AOD-100 related measurements and radiance measurements of CE318 #1098 and #0971 were calibrated via inter-comparison with a master instrument using Langley method and using a vicarious/transfer calibration method, respectively (Li et al., 2008;.
The intensive measurements of four radiometers in this campaign aim to verify the consistencies of calibration coefficients.
The volume aerosol parameters were combined with four radiometer measurements following SONET level 1.5 data criteria (Li et al., 2018). Observations from the CE318-T #1141 during the DAO-K campaign also joined in the AERONET dataset. 105 The consistency of the products following the AERONET and SONET retrieval frameworks has been validated by Li et al. (2018). In the retrieval frameworks, the dust particles were considered as non-spherical particles, which were modelled by randomly oriented spheroids (Dubovik et al., 2006).
A solar radiation monitoring station, consisting of an EKO MS-57 pyrheliometer and two MS-80 pyranometers, was used for measuring the direct, diffuse and total solar irradiances in the range of 0.28~3.0 µm. from which the vertical distribution of multiple optical and physical properties of dust aerosols (Veselovskii et al., 2016(Veselovskii et al., , 2018Hu et al., 2019) can be obtained. The backscattering coefficient profile at 355 nm wavelength was applied in this study to distinguish the two-layer structure of dust. PM2.5 mass concentration was measured by a continuous particulate monitor. The hourly PM10 mass concentrations in Kashi were collected from the routine measurements operated by China National Environmental Monitoring Center (CNEMC). The cloud-cover automatic observation instrument equipped with two wide-120 dynamic full-sky visible and infrared imagers, detected cloud amount and distribution in the sky during day and night with 10 min (or less than 10 min) resolution. Atmospheric profiles during the campaign were collected from sounding balloon measurements operated by Kashi regional meteorological bureau.
In addition to the ground-based observations, the satellite products of Moderate resolution imaging spectroradiometer (MODIS)/Terra+Aqua were adopted to derive the diurnal-change of the surface albedo (Schaaf and Wang, 2015). Ozone 125 profiles obtained by the Ozone Monitoring Instrument (OMI)/Aura (Bhartia et al., 1996) were applied in the RT model input.
Considering different durations of various measurements, in this study we calculated and discussed the ASRF from 2 to 25 April 2019, when simultaneous measurements are available.   suggesting that the heavy aerosol outbreaks at Kashi were dominated by dust particles. As a qualitative indicator of aerosol particle size, the values of AE are always less than 1.0 during the DAO-K campaign, suggesting that aerosol particles around 140 the Taklimakan desert are mainly dominated by coarse particles (even for clear situations). This is consistent with the results https://doi.org/10.5194/acp-2020-60 Preprint. Discussion started: 17 February 2020 c Author(s) 2020. CC BY 4.0 License. obtained in a previous study (Fig. 4 in Li et al., 2018). Comparatively high values of AE (>0.4) can be observed on 7, 12, 19, and 23 April 2019, implying enhanced local anthropogenic pollution and relatively small particle enrichments for these days.

130
The average values of PM2.5 and PM10 mass concentrations corresponding to measuring times of CE318 are 0.08 and 0.25 mg m -3 during the campaign, respectively. The time series of PM2.5 and PM10 mass concentrations generally concur with that 145 of AOD. However, for some days such as 19 and 23 April 2019, relatively high PM2.5 corresponding to low AOD has been observed, also indicating the enhanced influences of anthropogenic pollutions in these cases. But for 7 and 12 April 2019, high AE values corresponding to low PM2.5 values could be down to the very low turbidity conditions. The errors in computations of AE significantly increase under low aerosol loading conditions (Kaskaoutis et al., 2007). The maximum PM10 concentration during the heavy dust storm episode from 24 to 25 April 2019 was up to 4 mg m -3 . However, only moderate values of PM10 150 are shown in Fig. 3 because there was no CE318 measurement around the peak time of dust outbreak. 3 Estimation of aerosol solar radiative forcing 155
where ASRFTOA, ASRFBOA and ASRFATM denote the direct aerosol solar radiative forcing at the TOA, BOA and in ATM, respectively. F a and F 0 indicate the net irradiances with and without aerosols, respectively. F  and F  separately represent the downward and upward irradiances. All the above quantities are in W m -2 . The radiative forcing efficiency is defined as the rate 165 at which the atmosphere is forced per unit of aerosol optical depth at 550 nm (Garcí a et al., 2008(Garcí a et al., , 2012: where ASRFE (in W m -2 τ550 -1 ) is the aerosol solar radiative forcing efficiency at the TOA, BOA, or in ATM. Since the effects of aerosol loading on radiative forcing have been eliminated, radiative forcing efficiency has unique advantage on evaluation of the direct radiative effects of different types of aerosols (Garcí a et al., 2008). 170

Radiative transfer model simulation
The focus of this study is to quantify of direct ASRF and ASRFE at the TOA, BOA, and in ATM under cloud-free sky conditions by the SBDART radiative transfer model with reliable satellite and ground-based observations during the DAO-K campaign as model inputs. SBDART is a radiative transfer software tool that has been widely adopted in atmospheric radiative energy balance studies (Ricchiazzi et al.,1998;Li et al., 2018). The discrete ordinate method is adopted in the code, which provides a 175 numerically stable algorithm to solve the equations of plane-parallel radiative transfer in a vertically inhomogeneous atmosphere (Ricchiazzi et al.,1998). The simulations cover the same wavelength range (i.e., 0.28~3.0 µm) with the pyranometer for the convenience of comparison. Simulations of the ASRF by RT model are susceptible to the input conditions including the aerosol properties, atmosphere profile (including ozone), and land surface albedo. They were specified based on the high-quality dataset obtained in the DAO-K campaign. 180

Setting of aerosol properties
The aerosol properties including AOD, SSA, AE, and g retrieved from the radiometer observations at four bands with the central wavelengths at 440, 675, 870, and 1020 nm. They were applied in the instantaneous radiative forcing and efficiency calculations at the corresponding observing time. The aerosol properties in the SW range are obtained by interpolation and extrapolation using parameters in the above mentioned four bands. For daily mean ASRF simulation, the averaged aerosol 185 parameters (i.e., AOD, SSA, AE, and g) obtained from the day-time radiometer observations were used as alternatives of the daily mean aerosol properties. The daily mean aerosol radiative forcing and efficiency were calculated by taking the average of the 24 instantaneous values on an hourly basis.

Setting of atmospheric profile
In addition to aerosol properties, atmospheric profiles of thermodynamic properties are important for ASRF calculation. The 190 vertical distributions of air pressure, temperature, water vapor, and ozone densities have obvious influences on the direct and diffuse solar irradiances at the BOA. The predefined atmospheric profiles in RT model (e.g., tropical, mid-latitude summer, mid-latitude winter, sub-arctic summer, sub-arctic profiles) are different from Kashi local conditions. Within the ASRF simulations, the predefined profiles have been replaced by the actual measurements conducted during the campaign. Vertical distributions of the atmospheric pressure, temperature, relative humidity can be obtained by atmospheric sounding twice a day 195 around 0:00 and 12:00 UTC at Kashi. The profiles of ozone density (in g m -3 ) was deduced from the OMI/Aura OMO3PR product (in DU) (Bhartia et al., 1996). Two atmospheric profiles were specified for each day. The profile closest to the simulated moment of ASRF was adopted for both of instantaneous and daily mean aerosol radiative forcing estimates.

Setting of land surface albedo
Land surface albedo (LSA) is another key factor to influence the radiation budget, mainly due to its significant impact on the 200 SW upward irradiance (Liang, 2004;Bierwirth et al., 2009;Tegen et al., 2009;Jä kel et al., 2013;Stapf et al., 2019). Shortwave land surface albedo SW  , also known as blue-sky albedo, can be calculated from the black-sky albedo (BSA) BSA SW  and white-sky albedo (WSA) WSA SW  weighted by the fraction of diffuse skylight radiation (Schaaf et al., 2002;Wang et al., 2015): where diffuse,SW f denotes the fraction of diffuse radiation in the solar spectral range. ss ( , )  specifies the incident solar 205 geometry (i.e., solar zenith angle and solar azimuthal angle).
The shortwave WSA and BSA are provided by the MODIS BRDF/Albedo Science Data Product MCD43A3, which is produced daily using 16 days of MODIS/Terra+Aqua data. MCD43A3 only delivers the surface albedo products at local solar noon. However, diurnal variations of LSA cannot be ignored, which has been demonstrated by several previous studies (Lewis and Barnsley, 1994;Lucht et al., 2000;Wang et al., 2015). There will be an obvious bias in estimating daily solar radiation 210 when simply using the local noon value as a surrogate of daily mean albedo (Wang et al., 2015). As for the weighting parameters of the RossThickLiSparseReciprocal BRDF model (i.e., isotropic, volumetric, and geometric), the changes within 16 days are subtle. Therefore, the daily three model weighting parameters over the SW band afforded by MCD43A1 are adopted to derived the WSA and BSA (the latter is as a function of incident solar direction) at different simulated moments of ASRF. The fraction of diffuse radiation can be calculated by the ratio of diffuse solar irradiance to total solar irradiance, which value of LSA in the morning (0.253) is greater than that in the afternoon (0.218), which has been supported by some other field observations (Minnis et al., 1997;Wang et al., 2015). The local noon albedo shows very low value. The daily mean albedo under the clear-sky condition (0.199) is appreciably greater than the local noon albedo (0.173). However, in the dustpolluted (almost the whole days of 9/4/2019 and 25/4/2019) and cloudy (the morning of 12/4/2019) sky conditions, the changes of LSA are not as severely as in the clear-sky conditions. Nevertheless, the local noon albedo still cannot reflect the effects of 225 aerosol and cloud variations on land surface albedo. Thus, diurnal-changed LSA and the daily mean albedo were adopted in the instantaneous and daily mean ASRF simulations, respectively. It is predictable that estimations of instantaneous and daily mean aerosol radiative forcing can be improved by considering diurnal variations of LSA instead of local noon albedo.

Forecast model
The Weather Research and Forecasting model with chemistry (WRF-Chem) model version 4.0 (Grell et al., 2005;Fast et al., 2006) was also used to simulate the ASRF at Kashi. The simulations were configured in a 9 km domain centered at Kashi site with 45×45 grid points and 41 vertical levels that extended from the surface to 50 hPa. The main physical options used for this 235 study included the Purdue Lin microphysics scheme, the unified Noah land surface model, the Yonsei University (YSU) scheme for planetary boundary layer meteorological conditions, and the Rapid Radiative Transfer Model for General Circulation Models (RRTMG) for SW and longwave radiation. The Carbon Bond Mechanism (CBMZ) was used for the Gasphase chemistry processes (Zaveri and Peters, 1999), which includes aqueous-phase chemistry. The aerosol chemistry was Aerosol types such as sulfate, methanesulfonate, nitrate, ammonium, black carbon, primary organic carbon, sodium, calcium, chloride, carbonate, aerosol liquid water, and other inorganic matter (e.g., trace metals and silica) are involved in the simulation.
Dust was simulated with the GOCART dust emission scheme (Ginoux et al., 2001). The dust particulates were aggregated into 245 the other inorganic matter component and were presented in the calculation of aerosol optical properties with anthropogenic aerosols.
Aerosol optical properties were calculated as a function of wavelength based on the Mie theory. The aerosol components within each size bin are assumed to be internally mixed. The mixing refractive indexes are the volume-weight average in refractive indexes of all aerosol components. Aerosol extinction and scattering coefficients and asymmetric factors for a 250 particulate per size bin are attained though searching a look-up Mie table by Chebyshev polynomial interpolation with the desired mixing refractive indexes and wet particulate radius. The value of particulate extinction coefficient multiplied with the particulate number concentration is volume extinction coefficient which is then multiplied with the height of layer to attain the layer AOD value. The sum of all layer AOD values over the four size bins is the columnar total AOD and is used for calculating AOD increments in the assimilation. 255

Assimilation system
Gridpoint Statistical Interpolation (GSI) 3DVAR assimilation system (Wu et al., 2002;Kleist et al., 2009) version 3.7 was applied to improve the simulated aerosols by assimilating the aerosol measurements at Kashi. This GSI version has been modified to assimilate the aerosol products (Liu et al., 2011;Schwartz et al., 2012). We assimilated our ground-based multi- into AOD per size bin using the WRF-Chem aerosol optical routine. The adjoint observation operators for AOD and particulate matter are given as ln( ) ln( ) where ni is aerosol number concentration in the ith size bin, and c are the observed AOD and particulate matter mass concentrations. As no aerosol extinction coefficient assimilated in this experiment, we assume the extinction coefficient per size bin is constant in grid at each model layer. Innovation of number concentration due to AOD constraint is therefore a proportion of change in model layer AOD to the observed columnar AOD, which is attained via iteration to minimize the cost 270 function. Innovation of number concentration due to the constraints of PM2.5 and PM10 are associated with the ratios (ri) of mass concentrations to number concentrations in a size bin estimated in the guess field, weighted by the proportion of the size number concentration, changing in the iteration, to the total particulate matter concentration.  Guenther et al., 2006). Two one-month WRF-280

Experimental setup
Chem simulations were performed for April 2019, discarding a one-week spin-up at the beginning of each simulation. The first one-month simulation was used for modelling background error covariance (BEC). The second one-month simulation was assimilated the observations of PM2.5, PM10 and AOD with GSI at 0:00, 6:00, 12:00 and 18:00 UTC with the assimilation window of ±3 h centered at the analysis times. The model was restarted from the meteorology and chemistry at analysis time and ran to the next analysis time. For the second one, each restart called the radiation routines twice which included and 285 excluded the aerosols, respectively, and the corresponding difference between the two calls in irradiances is aerosol radiative forcing.
A general way to model BEC is the National Meteorological Center (NMC) method that computes the statistical differences between two forecasts with different leading lengths (e.g. 12 and 24 h, or 24 and 48 h) but valid at the same time (Parrish and Derber, 1992). However, in our experiment, the WRF-Chem model strongly underestimated aerosol 290 concentrations and hence likely lowered the error magnitudes. For this reason, we assessed the standard deviations of the control variables over the entire one-month period at the four analysis hours (00, 06, 12, 18 h), respectively. Each standard https://doi.org/10.5194/acp-2020-60 Preprint. Discussion started: 17 February 2020 c Author(s) 2020. CC BY 4.0 License. deviation field was used for modelling a BEC repeatedly applied in the assimilations at the corresponding analysis hour. This approach represents the strong fluctuations of control variables as weather evolution during clear and dusty days. We expect fluctuations of aerosols over the different weathers are larger than the uncertainties due to different leading forecast lengths 295 and may give a better input field for modelling BEC. The observation errors for AOD and PM were 50% of natural logarithm of 0.01 and those errors of PM including measurement error and representative error depending on the grid size and the PM concentrations (Schwartz et al., 2012). The choice of 50% was determined by trying experimentally with different values, which can effectively assimilate measurements and will not excessively damage the model results.

Aerosol solar radiative forcing and efficiency
Results of instantaneous ASRF and ASRFE at TOA, in ATM, at BOA during the DAO-K campaign are given in Fig. 5. Both positive and negative values of ASRF, corresponding to warming and cooling effects respectively, can be found at top of the atmosphere (Fig. 5a). However, aerosols have only warming effects in the atmosphere (Fig. 5c) and cooling effects at the surface (Fig. 5e) during the DAO-K campaign. ASRFs at the TOA and BOA exhibit obvious negative correlations with AOD. 305 But positive correlation can be observed between ASRF in the atmosphere and AOD. From Fig. 5, it is evident that the dust aerosol has strong influences on the solar radiation budget. For the five aforementioned high aerosol loading episodes (see Sect. 2.2), the dust-dominant aerosols have stronger cooling effects at the TOA and BOA, and more significant warming effects in the atmosphere than other low aerosol loading situations. Moreover, the cooling effects at the BOA are more noticeable than which at the TOA, with the lowest values of -217 W m -2 and -119 W m -2 , respectively. 310 When ASRF is normalized by aerosol optical depth at 550 nm, the result of ASRFE is not sensitive to the aerosol loading.
However, a weak negative correlation between ASRFE and AE can also be observed at the BOA (Fig. 5f). That means, the ASRFE at the surface can roughly indicate the radiative forcing effects of different types of aerosols (Garcí a et al., 2008).
Relatively large fraction of small particles associated with large AE has stronger ASRFE for cooling the surface than other low AE situations. But for TOA and ATM, there are no obvious correlations between ASRFE and AE. Globally the cooling effect 315 of aerosols at Kashi is more efficient at the BOA than that at the TOA. That is in accordance with the results of ASRF. In comparison with ASRF, the variation of ASRFE is relatively moderate during the campaign. The strongest cooling effects on the TOA and BOA all appear in the episode of dust storm outbreak (i.e., 24 and 25 April 2019). But large dust particles in this case do not show extreme radiative forcing efficiency. Strong cooling efficiencies at the surface during the DAO-K campaign occur in the very clear cases with high AE on 7 April 2019 (Fig. 5f). 320 During the DAO-K campaign, the average values of daily mean ASRF at Kashi are -19±1 3 W m -2 at the TOA and -36±23 W m -2 at the BOA, which are slightly stronger than the multiyear average values at this site (i.e., -16 W m -2 at the TOA and -18 W m -2 at the BOA) obtained by the previous study (Li et al., 2018). These results are reasonable, since the campaign was performed in the dust-prone season and higher aerosol loading situations have stronger ASRF effects as discussed above.

Clear sky case
Instantaneous ASRF and ASRFE of the clear sky case on 7 April 2019 is depicted in Fig. 6. It was a typical cloud-free day at Kashi with AOD at 550 nm less than 0.22 for the whole day. As discussed above, the highest AE is observed on this day during the one-month campaign (see Fig. 3). Both cooling and warming effects of aerosols can be found at the top of atmosphere. 335 The cooling effects of ASRF are up to -19 W m -2 at the TOA and -48 W m -2 at the BOA, and the warming effect of ASRF is up to 50 W m -2 in the atmosphere. The corresponding extreme ASRFE values are -126, -236, and 263 W m -2 τ550 -1 , respectively It is apparent that the changes of ASRFE are more intense than the corresponding ASRF for the clear case.
https://doi.org/10.5194/acp-2020-60 Preprint. Discussion started: 17 February 2020 c Author(s) 2020. CC BY 4.0 License.  and -217 W m -2 , respectively), and stronger warming effect in ATM (ASRF up to 121 W m -2 ). However, we observe the extreme ASRFE values of -51, -99, and 55 W m -2 τ550 -1 at the TOA, BOA, and in ATM, respectively, indicating that the radiative forcing of dust is less efficient than that of the clear case. Moreover, the variations of ASRFE in the dust case are more moderate than which of ASRF. These are strikingly different from the clear case.

Two-layer dust case
On 9 April 2019 one extra layer suspending above the planetary boundary layer (PBL) was observed. Fig. 8 illustrates the observations of LILAS on 8 April. Lidar observations on 9 April 2019 are not shown because the lidar stopped working due to technical problems in the night of 8 April 2019. According to the backscattering coefficient profiles at 355 nm, the lower 355 layer and upper layer can be clearly identified. Lidar measurements indicate that aerosols in the layer above the PBL are probably dust particles because the derived high depolarization ratios agree with the values for dust. However, from lidar measurements we cannot draw unambiguous conclusion about the aerosol type in the PBL, because the incomplete overlap range of the lidar system is up to 800~1000 m. From Fig. 3, high AOD corresponding to low AE in the whole atmosphere and high PM2.5 and PM10 concentrations in the surface layer are exhibited from 8 to 9 April. It also suggests the complex pollutions 360 by two-layer dust particles during this pollution process. AOD at 550 nm on 9 April changes from 1.4 to 2.2 (Fig. 9). In consistent with the above heavy dust case, only cooling effects can be observed at the TOA and BOA, and only warming effect can be found in ATM for this case. The two layers of dust particles result in a TOA cooling up to -102 W m -2 , BOA cooling of up to -198 W m -2 , and atmosphere warming of up to 123 W m -2 . The absolute values of ASRF in this case are all less than those in the heavy dust case, suggesting the aerosols in the heavy dust case have more powerful radiative forcing effects. 365 Nevertheless, the extreme values of ASRFE are -62, -105, and 58 W m -2 τ550 -1 at the TOA, BOA and in ATM, respectively, indicating that dust particles have similar radiative forcing efficiencies in the two-layer and heavy dust cases.  ASRF. The differences in the results of total downward irradiance (TDI), total upward irradiance (TUI), and ASRF at the TOA and BOA simulated with pre-defined midlatitude winter profile and user-specified profiles, and simulated with local noon surface albedo and instantaneous surface albedo are given, respectively. According to Fig. 10a, different settings of profiles 375 have no influence on the TDI at the TOA. For the TUI, the absolute differences are less than 9 W m -2 . However, the atmospheric profile has significant impacts on both the TDI and TUI at the surface. The influences on TDI are generally stronger than which on TUI. The maximum absolute difference is up to 138 W m -2 (Fig. 10c). For ASRF at the TOA, the effects of atmospheric profiles are less than 5 W m -2 . But the serious influences can up to 103 W m -2 on ASRF at the BOA (Fig. 10e). The average effect of different profiles on ASRF is 0.8 W m -2 at the TOA, which is quite small in comparison with the average values of 380 daily ASRF (-19 W m -2 ). However, the average difference of 13 W m -2 for ASRF affected by atmospheric profiles cannot be ignored relative to the average ASRF (-36 W m -2 ) at the BOA. As a result, the cooling effects of aerosol radiative forcing will be significantly underestimated at the BOA simulated with the pre-defined midlatitude winter profile instead of the userspecified Kashi profiles.
Like atmospheric profile, different settings of LSA have also no influence on TDI at the TOA (Fig. 10b). They have small 385 effects on TDI at the BOA (absolute difference less than 3 W m -2 ), but obvious impacts on TUI at the TOA and BOA (absolute difference up to 22 W m -2 ) (Fig. 10 b,d). From Fig. 4, the local noon albedo is often less than the daily mean albedo. Especially for the clear day, the minimum of LSA occurs around the local noon. Then the TUI at the TOA and BOA will generally be underestimated by using the local noon albedo instead of instantaneous surface albedo in the simulations. But for ASRF (Fig.   10f), two LSA settings lead to moderate impacts at the TOA and BOA with average difference of 1.8 and 1.7 W m -2 , 390 respectively. Therefore, simulations using the local noon albedo trend to overestimate the cooling effects of the aerosol radiative forcing both at the TOA and BOA.

Comparisons and validation
In order to evaluate the results of aerosol radiative forcing comprehensively, we compare the radiative transfer simulations with the AERONET operational products and the WRF-Chem simulations. The downward irradiance at the surface directly measured by high-precision solar radiation monitoring station during the DAO-K campaign are further adopted to validate the 405 RT model and WRF-Chem model simulations.

Comparison between radiative transfer simulations and AERONET results
Aerosol radiative forcing at the TOA and BOA are operational products provided routinely by AERONET. Measurements of the CE318-T #1141 during the DAO-K campaign have been processed by AERONET. Therefore, we can compare the ASRF product from AERONET with our simulations. It should be noted that AERONET adopts different definition of ASRF that 410 only taking the downward irradiance at the BOA and the upward irradiance at the TOA into consideration. Omitting the downward irradiances with and without aerosols in the AERONET definition won't make much difference of ASRF at the TOA. But for ASRF at the BOA, neglecting the upward irradiance with and without aerosols in the AERONET definition will lead to obvious difference. Some existing studies have executed this kind of comparison (Garcí a et al., 2008;Garcí a et al., 2012;Bi et al., 2014) and reported that AERONET trends to overestimate aerosol ASRF at the BOA (Garcí a et al., 2012). 415 Fig. 11 presents the correlations of instantaneous aerosol ASRF between the RT model simulations and the AERONET products. It is obvious that there are linear relationships between our RT simulations and the AERONET results with R 2 up to 0.98 and 0.99 at the TOA and BOA, respectively. Two ASRF results at the TOA show good consistency with a slope of 1.01, even though the calculated SW ranges are not an exact match (i.e., 0.28~3.0 µm for this study, and 0.2~4.0 µm for AERONET).
But for BOA, the AERONET products are obvious greater than the corresponding RT model simulations (with a slope of 1.24), 420 which is in agreement with the conclusion of the previous study (Garcí a et al., 2012).

Comparison between radiative transfer model and WRF-Chem model simulations 425
Fig. 12 compares the assimilated aerosols to the observations. Evidently, the assimilation greatly improves the particulate matter concentrations and show reasonable variations in accordance with the dust episodes. However, two disadvantages are noticeable. One is the assimilation fails to reproduce the extremely high PM2.5 and PM10 on 24~25 April 2019, because the BEC is not specific for the model error in the strong dust storm. A better model result for the specific dust storm requires improving the model capability of simulating dust emission and the transport of dust particulates besides data assimilation. 430 Another is the assimilated AOD indeed increases but not well approaches the observations. The reason is that we only https://doi.org/10.5194/acp-2020-60 Preprint. Discussion started: 17 February 2020 c Author(s) 2020. CC BY 4.0 License. assimilated AOD by assuming the invariable extinction coefficients. Hence, this low bias in AOD cannot be eliminated by choosing a scaling factor smaller than 50% in the observation error for it will damage the surface-layer particulate results. As a result, we give a priority to the high quality of the surface-layer aerosol assimilation, and the aerosol optical depth in the assimilated WRF-Chem results is underestimated, which should be kept in mind when comparing the WRF-Chem results with 435 the RT model simulations.  and WRF-Chem models. Two results show similar variation patterns. However, it is notable that the WRF-Chem results are significantly greater than which of RT simulations in dust-polluted cases on 9, 24, and 25 April 2019. According to RT simulations, the strongest radiative forcing occurred on 25 April 2019. However, the most significant ASRF of WRF-Chem simulation is found on 24 April 2019 followed by 25 April 2019. As mentioned above, heavy dust storms broke out on these two days during the DAO-K campaign. The extreme values of daily mean ASRF calculated by RT model are -46 W m -2 at the TOA, 48 W m -2 in ATM, and -94 W m -2 at the BOA, but with respect to the WRF-Chem simulations, the corresponding values are -78, 101, and -178 W m -2 , respectively. The significant differences between the two kinds of simulated results in dust cases should be further evaluated.

Validation by ground-based irradiance measurements
Fig. 14 directly compares the RT and WRF-Chem simulated downward irradiances at surface with the ground-based measurements under three different sky conditions (i.e., clear case, heavy dust case, and two-layer dust case). The RT simulations of total, direct, and diffuse downward irradiances in the three situations agree well with high-precision 455 measurements of pyrheliometer and pyranometers. The percent differences of RT-simulated total irradiance with respect to the measurements are only 0.03% for the clear case, -2.67% for the heavy dust case, and -0.43% for the two-layer dust case.
Except for the heavy dust case, they are within the pyranometer measurement uncertainties (0.66%). As for the WRF-Chem simulations, the total irradiances in the clear sky case are consistent with RT simulations and measurements (Fig. 14a). But for the direct irradiances, there are obvious differences between the WRF-Chem simulations and the corresponding measurements 460 https://doi.org/10.5194/acp-2020-60 Preprint. Discussion started: 17 February 2020 c Author(s) 2020. CC BY 4.0 License. (Fig. 14b). Moreover, the WRF-Chem simulated diffuse irradiances in the clear case (Fig. 14c), the total, direct, and diffuse irradiances in the heavy dust and two-layer dust cases (Fig. 14d~i) are significantly distinct from the measurements and RT simulations.
One of the most noticeable features in the curves of WRF-Chem results is the sudden jump around 6:00 UTC, which can be attributed to data assimilations restarted at 6:00 UTC and ran to the next analysis time 12:00 UTC. The WRF-Chem results 465 are greatly improved after 6:00 UTC in the dust-polluted cases. It is evident that data assimilations at 6:00 UTC can ameliorate the WRF-Chem simulations in dust cases, but the correction effects are still limited. So, the problems of the WRF-Chem simulation have not yet been fully resolved by the assimilations of aerosol optical depth and particulate matter concentrations.
This conclusion is in accordance with Figs. 12 and 13. Our measurements have proved that the simulations of RT model are reliable in both of clear and high aerosol loading situations. The WRF-Chem model preforms better in clear sky than in the 470 dust-polluted conditions. There is still room for improving the WRF-Chem simulation of dust aerosol radiative forcing.

Summary and conclusions
Dust aerosol particles play an important role in local and global climate changes by influencing the radiation budget through scattering and absorbing processes, especially for the region close to dust sources such as deserts. The complicated scattering and absorption characteristics of dust particles make it challenging to estimate their direct radiative forcing. Therefore, the 480 Dust Aerosol Observation-Kashi (DAO-K) campaign was designed and preformed near the Taklimakan desert that is a substantial and stable source of Asian dust aerosol particles. For almost one month, comprehensive observations of aerosol properties (i.e., aerosol optical depth, Ångstrӧm exponent, single scattering albedo, and asymmetry factor), atmospheric profiles (including ozone profiles), and land surface properties were obtained by a variety of ground-based and satellite apparatus in the dust borne season, and were applied in aerosol solar radiative forcing analysis using the SBDART radiative 485 transfer model simulations. In addition to high-quality dataset of volume aerosol properties satisfying the AERONET and SONET level 1.5 data criteria, the daily specified atmospheric profiles and diurnal variations of land surface albedo were also considered in detail in the simulations. The simulated results show that the average values of daily mean ASRF at Kashi are -19 W m -2 at the TOA and -36 W m -2 at the BOA during the DAO-K campaign. The dust-dominant aerosols have stronger cooling effects at both the top and bottom of atmosphere, and more significant warming effects in the atmosphere than other 490 low aerosol loading situations. Nevertheless, the radiative forcing efficiencies in dust-polluted cases exhibit lower than those in clear-sky conditions. The average influences of different profiles on ASRF are small at the TOA (0.8 W m -2 ) but remarkable at the BOA (13 W m -2 ). The cooling effects of aerosol radiative forcing at the BOA will be significantly underestimated by simulations with the pre-defined midlatitude winter profile instead of the user-specified Kashi profiles. Simulations using the local noon albedo trend to overestimate the cooling effects of the aerosol radiative forcing both at the TOA and BOA. Different 495 land surface albedo settings lead to moderate impacts on ASRF with average effects of 1.8 W m -2 at the TOA and 1.7 W m -2 at the BOA.
By assimilating the multi-wavelength volume AOD and the surface-layer PM2.5 and PM10 mass concentrations, the aerosol solar radiative forcing was also simulated for the time period of DAO-K field campaign using the WRF-Chem model. The measurements of downward irradiances at the surface were applied in evaluating the two kinds of simulations. By comparison 500 of the daily mean ASRF, two results present similar variation patterns. However, the WRF-Chem results are significantly stronger than the RT simulations in dust-polluted cases. For the heavy dust episode, the percent difference of daily mean ASRF between RT model and WRF-Chem model simulations are greater than 50% at the TOA, BOA, and in ATM. The direct, diffuse (and the sum of both) downward irradiances simulated by RT model in the clear sky, heavy dust, two-layer dust conditions are all in good agreement with ground-based measurements. As for the WRF-Chem simulations, the total irradiances 505 in the clear sky case are consistent with RT calculations and measurements. But the direct, diffuse, and total irradiances simulated by WRF-Chem significantly deviate from measurements in the dust-polluted situations. Data assimilations can https://doi.org/10.5194/acp-2020-60 Preprint. Discussion started: 17 February 2020 c Author(s) 2020. CC BY 4.0 License. obvious improve the WRF-Chem simulations in dust cases, but the correction effects are still limited. Based on these findings it is concluded that the SBDART radiative transfer model provides credible estimates of dust particle solar radiative forcing if supplied with reliable model inputs, but the WRF-Chem model is prone to overestimate the radiative forcing effects of dust 510 aerosols. Considering the actual measured atmospheric profiles and diurnal variations of land surface albedo can improve the radiative transfer model simulations. Optimizations of dust emission scheme, background error setting of dust assimilation system, dust parameterization including nonsphericity, are proposed as the promising approaches to improve the WRF-Chem simulations of dust radiative forcing. We would like to emphasize, however, that the comparison is only conducted at one site and in a limited time period in this study. Future research on this topic should include a systematic evaluation of RT and WRF-515 Chem model simulations on larger space and time scales.