1. Introduction
Mineral (desert) dust emission from the surface is the result of wind erosion in dry areas with limited vegetation cover (Prospero et al. 2002). The evidence of profound variability in the dust cycle in response to climatic changes is imprinted in natural archives (Lambert et al. 2008; Kohfeld and Harrison 2001). Climatic (Ginoux et al. 2012; Tegen and Fung 1995), biogeochemical (Jickells et al. 2005), and economic and public health dust-related issues (Pérez García-Pando et al. 2014) call for attention on the future of the dust cycle. Because of the large fluctuations seen in the paleoclimate records (Lambert et al. 2008; Kohfeld and Harrison 2001) the representation of the dust cycle is now part of the strategy of phase 6 of the Coupled Model Intercomparison Project (CMIP6) and phase 4 of the Paleoclimate Modeling Intercomparison Project (PMIP4) (Eyring et al. 2016; Kageyama et al. 2017, 2018; Otto-Bliesner et al. 2017). In the atmosphere, dust scatters and absorbs solar and terrestrial radiation (direct effect) (Sokolik and Toon 1996; Ginoux et al. 2001) and impacts cloud albedo and lifetime (indirect effects) (Atkinson et al. 2013). By carrying elements such as phosphorus and iron, dust also has an impact on the global biogeochemical cycles and ultimately on the carbon cycle (Mahowald et al. 2017). Once deposited, dust can also modify snow and ice albedos (Krinner et al. 2006).
Here we focus on the direct radiative effects (DRE) of dust on climate. Similar to other aerosols, dust impacts on climate are characterized by high uncertainty, partly because of the spatial and temporal variability of their emissions (Boucher et al. 2013). In addition, size (Mahowald et al. 2014; Tegen and Lacis 1996), composition (Perlwitz et al. 2015b; Scanza et al. 2015), and potentially shape (Kalashnikova and Sokolik 2004; Potenza et al. 2016) variability in dust aerosol all contribute to determining the longwave (LW) and shortwave (SW) DRE. The net effect on the radiation budget is thus very sensitive to how these features are modeled (Sokolik and Toon 1996; Albani et al. 2014; Miller et al. 2004), so that even the sign of these effects is not fully constrained (Boucher et al. 2013; Kok et al. 2017). In general, three parameters will determine dust direct effects in a model: dust physical properties (intrinsic optical properties and particle size distributions), atmospheric dust load, and the surface albedo (Yu et al. 2006). In addition, simulated cloud cover and parameterizations of boundary layer mixing will mediate how dust DRE will translate into surface climate impacts (Miller et al. 2014).
Here we use the Community Earth System Model (CESM) to estimate the dust direct impacts on climate, in terms of DRE as well as impacts on surface temperatures and precipitation. Previous studies based on global climate models have looked at the surface climate effects of dust for the present day (e.g., Miller and Tegen 1998; Yoshioka et al. 2007) or in rare cases the Last Glacial Maximum (Mahowald et al. 2006), but not across different paleoclimate conditions. A recent study analyzed the potential climate impacts of dust in the tropics during the mid-Holocene (Pausata et al. 2016). For the first time we consider different climate regimes of the present and the past, and compare a set of simulations with the dust radiative impacts enabled (DUST), contrasted with a parallel set of simulations without dust (NODUST). The climate states analyzed here include current climate (CUR), preindustrial conditions (PI; AD 1850), the mid Holocene (MH) 6000 years before the present (6 ka BP), and the Last Glacial Maximum (LGM; 21 ka BP). These periods encompass large fluctuations in the dust cycle, which may provide boundaries for possible future variations. In particular during the MH global dust emissions were at relatively low levels mostly due to the reduced emissions from the Sahara and Sahel during the so-called African humid period, characterized by an intensification of the West African monsoon (deMenocal et al. 2000). At the other extreme, dust emissions during the LGM were more than doubled compared to late Holocene levels; in this case a major role was played by increased emissions from mid- and high-latitude sources, in response to increased wind gustiness and aridity (McGee et al. 2010), expanded desert source areas (Mahowald et al. 1999), and/or the enhanced availability of fine-grained erodible material produced by glacial erosion and discharged on glacial outwash plains (Maher et al. 2010; Bullard 2013).
The dust model embedded in the CESM (Mahowald et al. 2006) was compared with a wide set of present-day observations from different platforms and regions, in order to realistically constrain the magnitude of dust load, surface concentration, deposition, optical properties, and particle size distributions (Albani et al. 2014). The magnitude of emissions for past climate regimes was constrained based on compilations of paleodust mass accumulation rates and particle size distributions (Albani et al. 2014, 2015, 2016). This particular strategy was explicitly aimed at having the most realistic representation of the mass budgets of dust, in order to focus on studying dust impacts onto climate under very different climate and dust conditions.
2. Methods
The CESM is a fully coupled global Earth system model that represents the dynamics and exchanges of energy, momentum, and mass between the land, the atmosphere, the oceans, the cryosphere, and the biosphere (Gent et al. 2011). The dust model embedded in the CESM represents dust emission, three-dimensional transport, dry and wet deposition, and interactions with both SW and LW radiation (Mahowald et al. 2006). Dust is emitted as a function of wind speed, soil moisture, vegetation, and snow cover, and uses a geomorphic source function to account for the spatially variable availability of erodible material (Zender et al. 2003a,b).
For this study we used the CESM version 1.0.5. The atmospheric component is the Community Atmosphere Model version 4 associated with a bulk atmospheric model (CAM4-BAM), which provides a sectional representation of the aerosols organized in four size bins (diameters of 0.1–1.0, 1.0–2.5, 2.5–5.0, and 5.0–10.0 μm). The dust emission flux is calculated based on a saltation and sandblasting scheme (Zender et al. 2003a) and is partitioned into the four aerosol bins with a fixed size distribution (Kok 2011). The model simulations are performed with the fully coupled setup, including a dynamic ocean model. Vegetation is prescribed according to PMIP3 (Brady et al. 2013), but the land component includes prognostic vegetation phenology coupled to the carbon and nitrogen cycles (Lawrence et al. 2011). Dust feedbacks include absorption and scattering of SW radiation, absorption of LW radiation, and changing snow albedo (Neale et al. 2010; Flanner et al. 2007; Yoshioka et al. 2007). The cloud parameterizations in CAM4 use a bulk microphysics scheme (Rasch and Kristjánsson 1998) and do not represent aerosol indirect effects. The planetary boundary layer (PBL) is represented by a nonlocal transport scheme, optimized for simulation of dry convective and nocturnal PBL over land (Neale et al. 2010). Further details on the current setup and parameterization suite are fully described in Albani et al. (2014).
Albani et al. (2014) evaluated the representation of the dust cycle in the CESM by comparing simulations for current climate against extensive compilations of in situ and remote sensing measurements, by looking at the climatological annual average and annual cycles of dust surface concentrations, column-integrated properties such as dust aerosol optical depth (AOD), and deposition fluxes to the surface. In addition, we evaluated the size distributions and clear-sky DRE efficiency of the model against the limited observations/estimates available (Albani et al. 2014). For past climate regimes, we started from the end of PMIP3 simulations with the CESM (Brady et al. 2013), and we accounted for the influence of vegetation changes on dust emissions by scaling the erodibility of individual grid cells in proportion to the vegetation fraction, as simulated by the biogeochemistry–biogeography model (BIOME4) (Albani et al. 2014, 2015; Kaplan et al. 2003).
For both current climate and paleoclimate regimes we focused on providing the most accurate possible representation of the magnitude of the dust cycle (i.e., the amount of dust) for climate equilibrium conditions; in order to do so, we further refined the emissions by applying a set of scale factors at the macro area/continental scale in order to obtain a better fit to the observational estimates of deposition fluxes from paleoclimate archives in each climate regime, as well as to be consistent with indications of dust provenance (Albani et al. 2014; Mahowald et al. 2006; Albani et al. 2015).
A full description of the simulations of dust cycle in the CESM and evaluation against observational data for current climate (CUR; AD 2000), Last Glacial Maximum ~21 000 years before present (LGM; 21 ka BP), mid-Holocene (MH; 6 ka BP), and preindustrial (PI; AD 1850) is provided elsewhere (Albani et al. 2014, 2015, 2016). For this study we integrate the existing simulations (referred to as DUST cases) with the new corresponding “base” simulations without dust (NODUST cases). All simulations were run for 50 years starting from climate equilibrium conditions, and our analyses are based on the last 30 years.
The length of our simulations is well within the range of recommendations for model intercomparison experiments. For instance PMIP4/CMIP6 (Kageyama et al. 2018) experiments for MH and LGM request 100 years after spinup. In our case, we started from LGM and MH spun-up models already, and only added dust on top of that, which is a relatively small change in radiative forcing compared with the change due to orography and carbon dioxide for the LGM, for example. Therefore our case is arguably more similar to the AerChemMIP experimental design for CMIP6 piClim-2xdust equilibrium experiment, for which the requirement is 30 years for simulation (Collins et al. 2017). Nonetheless, because our setup includes a dynamic ocean model, ocean feedbacks may not be fully captured in our experiments due to the length of the simulations. To test for the sensitivity to the length of the simulations, we extended the two LGM simulations (DUST, NODUST) to 200 years.
A further experiment was performed to test the sensitivity of dust properties with respect to its interaction with radiation. The sensitivity test uses the release version of CAM4 dust (Neale et al. 2010), which is characterized by finer size distributions (Mahowald et al. 2006; Yoshioka et al. 2007), more absorbing SW optical properties based on OPAC (Hess et al. 1998), and the absence of dust–LW interactions. A thorough comparison with observations and the “new” dust used for the main simulations of this paper is described in Albani et al. (2014). The sensitivity test is based on a simulation of 30 years with the same initial conditions as the current climate simulation (CUR), and the last 10 years were analyzed.
The instantaneous DRE (iDRE) is a diagnostic variable calculated at each model time step, as a difference of the net all-sky (i.e., including with clouds) radiative fluxes with all aerosol species, and with all except for dust. The effective direct radiative perturbation (eDRP) is calculated offline from the diagnostic variables on the monthly history files, by the difference of the net all-sky radiative fluxes between the simulations with and without dust, and also reflects adjustments of the climate system to dust forcing (Sherwood et al. 2015), on a scale of 30 years in this case.
Throughout the manuscript, we show either the annual (ANN) or seasonal averages [e.g., Northern Hemisphere summer (JJA)] depending on the context and the specific aspect being discussed. For instance, when discussing the summer monsoon we focus on JJA. On the other hand if we are giving an overview of the global dust impacts, ANN appears to be preferable. Complementary information is available in the online supplemental material.
3. Results
a. Overview
In line with the general understanding that we have from global paleodust compilations (Kohfeld and Harrison 2001; Albani et al. 2015; Maher et al. 2010), the simulated dust cycle is characterized by largest dust emission and load during the LGM (more than twice the late Holocene levels), with an expansion of mid- and high-latitude sources (Table 1, Figs. 1a,d). The deglaciation was characterized by reduced dust emissions (Lambert et al. 2008), and the MH in particular was characterized by a minimum in the global dust load, with highly reduced emissions from North Africa corresponding to the “Green Sahara” phase (deMenocal et al. 2000; McGee et al. 2013); the late Holocene was characterized by a return to increased dust emissions from North Africa (McGee et al. 2013; Albani et al. 2015). Our global dust emissions and load are within the range of other models for CUR (Huneeus et al. 2011) and LGM conditions (Albani et al. 2018). For the MH our simulation (Albani et al. 2015) is still the only attempt with a high-complexity global model (i.e., excluding studies based on intermediate complexity models, or only focused on a specific area without a full evaluation of the dust cycle); it is expected that PMIP4 dust experiments will provide a broader context to analyze simulations of the global dust cycle at 6 ka BP (Otto-Bliesner et al. 2017).
Global budgets for the four climate regimes analyzed in this paper: current climate (CUR), preindustrial (PI), mid-Holocene (MH), and the Last Glacial Maximum (LGM); iDRE denotes the instantaneous DRE and eDRP is the effective DRE (see section 2). For temperature and precipitation, statistically significant anomalies in the global averages (two-tailed t test; alpha = 0.10) are marked with an asterisk (*). Absolute eDRP indicates the global average of the absolute values of eDRP at TOA. Similarly, absolute temperature (precipitation) anomaly is calculated as the global average of the absolute values of the differences in temperature (precipitation) between the DUST and NODUST cases.
Zonal average plots of (a) dust load (g m−2), (b) dust net surface iDRE (W m−2), and (c) dust net TOA iDRE (W m−2). Different climate regimes are highlighted by different colors. In (a)–(c) the vertical solid gray lines highlight the zero value. In all panels, the horizontal dotted gray lines mark reference latitudes highlighted throughout the manuscript (limits in °N: −50, −25, 0, 25, 50). Also shown are maps of LGM (d) dust load (g m−2), (e) dust net TOA iDRE (W m−2), and (f) statistically significant (two-tailed FDR test; alpha = 0.20) dust net TOA eDRP (W m−2). The instantaneous radiative forcing (iDRE) is a diagnostic variable reflecting potential perturbations to the atmospheric radiation budget, while the effective DRE (eDRP) reflects the net impacts of dust on the climate system, including forcing on scale of 30 years.
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
Zonal average plots of (a) dust load (g m−2), (b) dust net surface iDRE (W m−2), and (c) dust net TOA iDRE (W m−2). Different climate regimes are highlighted by different colors. In (a)–(c) the vertical solid gray lines highlight the zero value. In all panels, the horizontal dotted gray lines mark reference latitudes highlighted throughout the manuscript (limits in °N: −50, −25, 0, 25, 50). Also shown are maps of LGM (d) dust load (g m−2), (e) dust net TOA iDRE (W m−2), and (f) statistically significant (two-tailed FDR test; alpha = 0.20) dust net TOA eDRP (W m−2). The instantaneous radiative forcing (iDRE) is a diagnostic variable reflecting potential perturbations to the atmospheric radiation budget, while the effective DRE (eDRP) reflects the net impacts of dust on the climate system, including forcing on scale of 30 years.
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
Zonal average plots of (a) dust load (g m−2), (b) dust net surface iDRE (W m−2), and (c) dust net TOA iDRE (W m−2). Different climate regimes are highlighted by different colors. In (a)–(c) the vertical solid gray lines highlight the zero value. In all panels, the horizontal dotted gray lines mark reference latitudes highlighted throughout the manuscript (limits in °N: −50, −25, 0, 25, 50). Also shown are maps of LGM (d) dust load (g m−2), (e) dust net TOA iDRE (W m−2), and (f) statistically significant (two-tailed FDR test; alpha = 0.20) dust net TOA eDRP (W m−2). The instantaneous radiative forcing (iDRE) is a diagnostic variable reflecting potential perturbations to the atmospheric radiation budget, while the effective DRE (eDRP) reflects the net impacts of dust on the climate system, including forcing on scale of 30 years.
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
The wide range of estimates in dust load among models is reflected in the corresponding estimates of dust DRE (Boucher et al. 2013; Albani et al. 2018)—because dust interacts with both SW and LW radiation, it is important that both are parameterized in current dust models (Albani et al. 2014; Sokolik et al. 1998). Because of the sensitivity of the dust DRE to size and composition assumptions (Perlwitz et al. 2001), it is vital that available estimates of DRE by dust based on observations are compared to the corresponding variables in model experiments (e.g., Albani et al. 2014). We now take as an example the results of the LGM simulation, when dust loads and the associated DRE are more pronounced. In our simulations the dust net (SW+LW) iDRE has a negative sign that scales almost linearly with dust loading across different climates, with a global surface negative forcing only in part compensated by a positive forcing through the atmospheric column, resulting in a negative top of the atmosphere (TOA) iDRE (Table 1). The zonal distribution of the forcing also mirrors the distribution of dust load (Figs. 1b,c; see also Fig. S1 in the online supplemental material). The global precipitation is consistently reduced by dust, even if this reduction is very small; on the other hand global temperature anomalies are generally insignificant and inconsistent in sign (Table 1). This is actually in line with the almost zero global eDRP global budget across climates (Table 1).
These small global averages result from partial compensation between large regional DRE anomalies of opposing sign in different regions (Figs. 1e,f). If we consider the global average in terms of the absolute values of net TOA eDRP instead, the resulting budgets are quite remarkable, ranging between |0.23| and |0.49| W m−2 (Table 1). Because the local impacts will be proportional to these changes in radiation, we regard this as an important metric to summarize the direct impacts of dust on climate, complementary to the eDRP itself. Based on our simulations, it scales linearly with dust load, yielding an absolute eDRP efficiency of ~0.013 W m−2 Tg−1dust. Similarly, if we compute the global average of the absolute value of the anomalies in precipitation and temperatures, we highlight the overall average perturbation of dust to these climate variables, which is nonnegligible, especially in the case of temperature anomalies (Table 1).
The maps of iDRE (Fig. 1e) and eDRP (Fig. 1f) show the same geographical pattern, but the absolute values of the regional forcing are slightly different; this feature in fact explains the differences in the global DRE budget in the two cases (Table 1). While dust generally causes a negative TOA DRE, positive values are found in correspondence of bright surfaces with elevated albedo (Patadia et al. 2009), such as some deserts and ice-covered regions (Figs. S1–S3).
The positive TOA DRE over the Arctic in our simulations results from both an atmospheric positive anomaly due to scattering of SW radiation (as shown by the diagnostic variable iDRE in Figs. 1b, 1c, and 1e) and a surface positive anomaly due to snow darkening (an effect that can be deduced from the eDRP variable in Fig. 1f).
The patchy pattern of SW TOA DRE over North Africa (Albani et al. 2014) reflects the differential response of the scattering dust particles over the varying surface albedo, with darker surfaces associated with the presence of vegetation in the Sahel and reliefs such as the Hoggar, Tibesti, and Atlas mountain ranges as opposed to the bright dry lowlands of the Saharan desert (Slingo et al. 2006; Patadia et al. 2009). A similar patchy pattern is also visible in the net surface DRE (Albani et al. 2014). It results from the opposing signals of SW and LW DRE predominating over each other depending on the local surface albedo and the localized effects of the major dust hotspots. In the major dust hotspots, the larger dust particles are intermittently emitted before settling at relatively short distances, as opposed to areas more downwind where the winnowing of particle size distributions to a finer aerosol is associated with a higher bulk value of dust single scattering albedo. On the other hand the overall climate impacts on surface climate will be largely driven by the net TOA DRE (Miller et al. 2014)—in this case, a positive net TOA DRE over the Sahara opposing a negative signal over the Sahel.
b. Dust direct impacts on surface climate
We can examine the spatially resolved dust impacts on surface climate by looking at the horizontal distribution of the annual average (absolute values) DUST and NODUST anomalies of surface temperature and total precipitation, where we focus only on significant anomalies, that is, where a two-tailed test for difference in the mean following the false detection rate (FDR) method (Wilks 2016) (test level alpha = 0.20) is significant (Figs. 2a,b; Figs. S4–S6; see also the supplemental material for details on this method, which provides more conservative results than the widespread t test). Our goal is to identify some general features in terms of surface climate, which can be attributable to dust, and are independent of the climate regime.
(a) Surface temperature (Temp.; K) and (b) precipitation (Precip.; mm day−1) annual (ANN) DUST − NODUST anomalies (anom.) for the LGM. Only significant anomalies are plotted (two-tailed FDR test, alpha = 0.20). Maps showing the consistent (c) temperature (Temp.) and (d) precipitation (Precip.) DUST − NODUST anomalies (anom.) across the four different climate regimes for the annual average (ANN). Only grid cells where all of the four DUST − NODUST anomalies agree on the sign are plotted, whereas the color scale highlights in how many cases, out of four, the anomalies were statistically significant (two-tailed FDR test, alpha = 0.20).
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
(a) Surface temperature (Temp.; K) and (b) precipitation (Precip.; mm day−1) annual (ANN) DUST − NODUST anomalies (anom.) for the LGM. Only significant anomalies are plotted (two-tailed FDR test, alpha = 0.20). Maps showing the consistent (c) temperature (Temp.) and (d) precipitation (Precip.) DUST − NODUST anomalies (anom.) across the four different climate regimes for the annual average (ANN). Only grid cells where all of the four DUST − NODUST anomalies agree on the sign are plotted, whereas the color scale highlights in how many cases, out of four, the anomalies were statistically significant (two-tailed FDR test, alpha = 0.20).
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
(a) Surface temperature (Temp.; K) and (b) precipitation (Precip.; mm day−1) annual (ANN) DUST − NODUST anomalies (anom.) for the LGM. Only significant anomalies are plotted (two-tailed FDR test, alpha = 0.20). Maps showing the consistent (c) temperature (Temp.) and (d) precipitation (Precip.) DUST − NODUST anomalies (anom.) across the four different climate regimes for the annual average (ANN). Only grid cells where all of the four DUST − NODUST anomalies agree on the sign are plotted, whereas the color scale highlights in how many cases, out of four, the anomalies were statistically significant (two-tailed FDR test, alpha = 0.20).
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
The example of LGM annual average (ANN) anomaly in surface temperature (Fig. 2a) shows a pattern that mirrors the TOA DRE (Figs. 1e,f), with a widespread surface warming over the Arctic, the Laurentide ice sheet, Tibet, and the deserts in North Africa and the Arabian Peninsula, and a cooling over the oceans downwind of the major dust sources. Positive precipitation anomalies are seen over the Sahel and part of the Arctic (Fig. 2b). Similar features emerge from the analysis of all other climates (Fig. S4).
If we look at the agreement across different climates, in terms of sign of the anomalies across all four climates, we see a consistent (“consensus”) signal showing distinctive, common warming patterns over the major desert belt in North Africa and Asia, as well as in parts of the Arctic, with cooling in the northern and equatorial Atlantic, the northern Indian Ocean, and the western equatorial Pacific (Fig. 2c), as the impact of dust is included in the different climate model simulations. The consistent increase in precipitation over the Sahara is also an emerging feature (Fig. 2d). The same analysis at the seasonal scale (Figs. S7 and S8) shows qualitatively the same patterns, more strongly marked during the NH summer, when dust impacts on climate are included in the model simulations. Note the significant northward shift of the summer ITCZ over North Africa and the North Atlantic in the LGM. In general, we note a tendency for the ITCZ to move toward the summer hemisphere in response to dust direct effects.
Except for the remarkable warming signal over the Arctic in the LGM only, robust signals in the other experiments seem more pronounced at low and middle latitudes, where there is much more dust (Figs. 2c,d). We mentioned that dust emissions in the model account for changes in vegetation in the different climates, based on preliminary offline simulations. However, for the physical climate in the model simulations described here, vegetation (plant functional types) was prescribed to be the same in all four climates, according to PMIP3 experimental design (Brady et al. 2013). Despite changes in foliage cover in response to different climate conditions, all four experiments essentially share the same vegetation maps and consequently a very similar distribution of surface albedo in the absence of snow (Albani et al. 2014). The possible implications will be discussed further in the next sections.
Dust net surface eDRP, as well as dust impacts on surface temperature and precipitation for key regions across all four climates analyzed here, are highlighted in Fig. 3.
(a) Dust impacts on climate across time, reposted as area-weighted annual average for the entire globe (gray lines), as well as for key regions: the Arctic (blue; latitudes > 60°N), Sahara (yellow; 15° to 25°N, −10° to +25°E), and Sahel (green; 10° to 15°N, −10° to +25°E). (b) Dust net surface iDRE and eDRP. (c) DUST − NODUST surface temperature anomalies. (d) Total precipitation. The statistical significance of the anomalies for the different variables are visible in the maps in Figs. 1 and 2 (see also Figs. S2 and S4).
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
(a) Dust impacts on climate across time, reposted as area-weighted annual average for the entire globe (gray lines), as well as for key regions: the Arctic (blue; latitudes > 60°N), Sahara (yellow; 15° to 25°N, −10° to +25°E), and Sahel (green; 10° to 15°N, −10° to +25°E). (b) Dust net surface iDRE and eDRP. (c) DUST − NODUST surface temperature anomalies. (d) Total precipitation. The statistical significance of the anomalies for the different variables are visible in the maps in Figs. 1 and 2 (see also Figs. S2 and S4).
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
(a) Dust impacts on climate across time, reposted as area-weighted annual average for the entire globe (gray lines), as well as for key regions: the Arctic (blue; latitudes > 60°N), Sahara (yellow; 15° to 25°N, −10° to +25°E), and Sahel (green; 10° to 15°N, −10° to +25°E). (b) Dust net surface iDRE and eDRP. (c) DUST − NODUST surface temperature anomalies. (d) Total precipitation. The statistical significance of the anomalies for the different variables are visible in the maps in Figs. 1 and 2 (see also Figs. S2 and S4).
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
c. Dust impacts on the Arctic climate
The warming signal in the Arctic is a consistent feature of dust direct radiative feedbacks on climate, whose intensity is mediated by the amount of atmospheric dust and extensive cryosphere cover. Both were at maximum during the LGM, when this warming effect is more evident. We analyze this effect in more detail, focusing on the LGM summer season, which dominates the ANN anomalies. The increase in surface temperatures is linked to the positive TOA DRE anomaly in response to the presence of dust, and it is particularly marked over the Siberian Arctic (Figs. 1f and 2a).
At surface, a strong positive eDRP is associated with reduced snow albedo by dust in Siberia and North America at the margins of the ice sheets (Figs. 4a,b). In the same areas, much increased sensible and latent heat fluxes (Figs. 4c,d) correspond to increased surface temperatures (Fig. 2a) and reduced snow cover (Fig. 4f), accompanied by significant increase in leaf area index (Fig. 4e). In future studies with a dynamical vegetation model, it would be interesting to investigate whether dust impacts on surface temperatures and snow cover may impact not only phenology, but more profoundly the vegetation composition itself.
Dust impacts on the surface (DUST − NODUST anomalies for LGM NH summer). (a) Dust net TOA eDRP. (b) Fractional relative anomaly in surface albedo. (c) Fractional relative anomaly in sensible heat fluxes. (d) Fractional relative anomaly in latent heat fluxes. (e) Fractional relative anomaly in the leaf area index. (f) Fractional relative anomaly in snow cover fraction. Only grid cells where all of the four DUST − NODUST anomalies agree on the sign are plotted, whereas the color scale highlights in how many cases, out of four, the anomalies were statistically significant (two-tailed FDR test, alpha = 0.20). Pink contours indicate ice sheets; gray contours indicate the present-day coastlines; black contours represent the LGM coastlines.
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
Dust impacts on the surface (DUST − NODUST anomalies for LGM NH summer). (a) Dust net TOA eDRP. (b) Fractional relative anomaly in surface albedo. (c) Fractional relative anomaly in sensible heat fluxes. (d) Fractional relative anomaly in latent heat fluxes. (e) Fractional relative anomaly in the leaf area index. (f) Fractional relative anomaly in snow cover fraction. Only grid cells where all of the four DUST − NODUST anomalies agree on the sign are plotted, whereas the color scale highlights in how many cases, out of four, the anomalies were statistically significant (two-tailed FDR test, alpha = 0.20). Pink contours indicate ice sheets; gray contours indicate the present-day coastlines; black contours represent the LGM coastlines.
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
Dust impacts on the surface (DUST − NODUST anomalies for LGM NH summer). (a) Dust net TOA eDRP. (b) Fractional relative anomaly in surface albedo. (c) Fractional relative anomaly in sensible heat fluxes. (d) Fractional relative anomaly in latent heat fluxes. (e) Fractional relative anomaly in the leaf area index. (f) Fractional relative anomaly in snow cover fraction. Only grid cells where all of the four DUST − NODUST anomalies agree on the sign are plotted, whereas the color scale highlights in how many cases, out of four, the anomalies were statistically significant (two-tailed FDR test, alpha = 0.20). Pink contours indicate ice sheets; gray contours indicate the present-day coastlines; black contours represent the LGM coastlines.
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
In addition to the strong surface and TOA eDRP, also the net atmospheric divergence term is positive (not shown). This is reflected in warming of the atmospheric column, as indicated by the positive anomalies in geopotential height at 700 hPa; the maxima of surface temperature anomalies appear to correspond to regions of increased southerly flow (Fig. 5b).
Focus over the Arctic region during LGM NH summer (JJA). (a) Dust TOA eDRP (colors) and surface temperature anomalies (black line contours, intervals at 0 and ± 1, 2, 4, 8, 16 K) for the DUST − NODUST case; the dotted pattern highlights negative temperature anomalies. (b) Anomalies in geopotential height (colors) and horizontal winds (arrows) at 700 hPa for the DUST − NODUST case. A reference wind vector is marked in green, and its velocity is reported in the top-right corner. The statistically significant anomalies of the different variables shown in this figure are visible in Fig. S9.
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
Focus over the Arctic region during LGM NH summer (JJA). (a) Dust TOA eDRP (colors) and surface temperature anomalies (black line contours, intervals at 0 and ± 1, 2, 4, 8, 16 K) for the DUST − NODUST case; the dotted pattern highlights negative temperature anomalies. (b) Anomalies in geopotential height (colors) and horizontal winds (arrows) at 700 hPa for the DUST − NODUST case. A reference wind vector is marked in green, and its velocity is reported in the top-right corner. The statistically significant anomalies of the different variables shown in this figure are visible in Fig. S9.
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
Focus over the Arctic region during LGM NH summer (JJA). (a) Dust TOA eDRP (colors) and surface temperature anomalies (black line contours, intervals at 0 and ± 1, 2, 4, 8, 16 K) for the DUST − NODUST case; the dotted pattern highlights negative temperature anomalies. (b) Anomalies in geopotential height (colors) and horizontal winds (arrows) at 700 hPa for the DUST − NODUST case. A reference wind vector is marked in green, and its velocity is reported in the top-right corner. The statistically significant anomalies of the different variables shown in this figure are visible in Fig. S9.
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
The warming anomaly caused by dust in western Siberia may contribute to explain why glacial–interglacial temperature anomalies for the warmest months, reconstructed from pollen data, are smaller in this region than at more southern latitudes, such as for instance southern Europe (Bartlein et al. 2011). The dust-induced warming in western Siberia may also have contributed to the smaller growth of an ice sheet in this region. Previous studies suggested that snow darkening in the LGM could prevent the formation of an ice sheet over northern Asia (Krinner et al. 2006; Ohgaito et al. 2018). In our analysis, we found that both atmospheric forcing (i.e., atmospheric absorption of dust over bright surfaces) and changes to snow albedo could significantly contribute to Arctic warming by dust.
d. Dust impacts on the West African monsoon
We now focus our attention on the North Atlantic and North Africa regions, in particular during the NH summer in the LGM, when dust impacts are strongest. The presence of dust causes a northward shift of the intertropical convergence zone (ITCZ), associated with an enhancement of the summer monsoon and increase in total precipitable water (Fig. 6c) (Lau et al. 2009).
Focus over the North African/North Atlantic region for the NH summer (JJA). (a) Vertical profile of zonal winds (NODUSTLGM) (contour lines; m s−1) and their (DUST − NODUST)LGM anomalies (colors), averaged across the region highlighted by the white box in (c); the horizontal green solid line marks the vertical level of wind anomalies displayed in (c). (b) Vertical profiles of the DUST − NODUST anomalies (anom.) in the air temperature (K) for the LGM, averaged across the region highlighted by the white box in (c). (c) (DUST − NODUST)LGM anomalies for total precipitable water (Precip.; colors; kg m−2) and winds (vectors); a reference wind vector is marked in green, and its velocity is reported in the top-left corner. (d) Sea level pressure (contour lines; hPa) for the NODUSTLGM case, and (DUST − NODUST)LGM anomalies (colors). The statistically significant anomalies of the different variables shown in this figure are visible in Fig. S10.
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
Focus over the North African/North Atlantic region for the NH summer (JJA). (a) Vertical profile of zonal winds (NODUSTLGM) (contour lines; m s−1) and their (DUST − NODUST)LGM anomalies (colors), averaged across the region highlighted by the white box in (c); the horizontal green solid line marks the vertical level of wind anomalies displayed in (c). (b) Vertical profiles of the DUST − NODUST anomalies (anom.) in the air temperature (K) for the LGM, averaged across the region highlighted by the white box in (c). (c) (DUST − NODUST)LGM anomalies for total precipitable water (Precip.; colors; kg m−2) and winds (vectors); a reference wind vector is marked in green, and its velocity is reported in the top-left corner. (d) Sea level pressure (contour lines; hPa) for the NODUSTLGM case, and (DUST − NODUST)LGM anomalies (colors). The statistically significant anomalies of the different variables shown in this figure are visible in Fig. S10.
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
Focus over the North African/North Atlantic region for the NH summer (JJA). (a) Vertical profile of zonal winds (NODUSTLGM) (contour lines; m s−1) and their (DUST − NODUST)LGM anomalies (colors), averaged across the region highlighted by the white box in (c); the horizontal green solid line marks the vertical level of wind anomalies displayed in (c). (b) Vertical profiles of the DUST − NODUST anomalies (anom.) in the air temperature (K) for the LGM, averaged across the region highlighted by the white box in (c). (c) (DUST − NODUST)LGM anomalies for total precipitable water (Precip.; colors; kg m−2) and winds (vectors); a reference wind vector is marked in green, and its velocity is reported in the top-left corner. (d) Sea level pressure (contour lines; hPa) for the NODUSTLGM case, and (DUST − NODUST)LGM anomalies (colors). The statistically significant anomalies of the different variables shown in this figure are visible in Fig. S10.
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
The net TOA positive DRE over the Sahara controls the temperature anomaly induced by dust at equilibrium (Miller et al. 2014). In our simulations this is consistent with an increase in surface temperature under the plume (Fig. 2) and the heating of the mid- and lower troposphere (Fig. 6b), which is associated with a low pressure anomaly near the surface that reinforces the Saharan heat low (Fig. 6d). The opposing dust TOA net DRE and associated temperature anomalies over the Sahara and Sahel (Figs. 1e,f and 2) are also reflected in the patterns of sea level pressure anomalies, indicating a strengthening of the low pressure system over the Sahara in JJA (Fig. 6d), which is also driving the 930-hPa winds anomalies (Fig. 6c), and a warm anomaly in the midtroposphere (Fig. 6b). The average (15°W–5°E) zonal winds patterns (Fig. 6a) show how the summer monsoonal (westerly) flow is intensified farther inland to the north by the presence of dust. In particular we see an enhanced West African westerly surface jet (10°–12°N over the eastern Atlantic and coastal West Africa), which can increase the Sahel precipitation by increasing moisture transport and reducing atmospheric stability (Grodsky et al. 2003; Pu and Cook 2010, 2012). The African easterly jet aloft (centered around 12°N and 600 hPa) is also displaced more to the north when dust radiative effects are included. These features, which are seen across all four climates (Figs. S11, S12), are in line with the general understanding of the links between the Saharan heat low system and the West African monsoon (Lavaysse et al. 2009; Schepanski et al. 2017). During the MH in particular, despite a relatively minor impact on precipitation in North Africa due to the relatively low dust load, there is still an enhancement of the West African westerly surface jet (Fig. S11), which was found to play a very important role in maintaining the wet condition in Africa (Patricola and Cook 2007).
The dust-induced surface temperature anomalies over North Africa (Figs. 2a,c) are in line with the anomalies in sensible and latent heat fluxes (Figs. 4c,d). In particular, over the Sahel a strong negative anomaly characterizes both sensible and latent heat fluxes, associated with the negative eDRP at the surface, and negative temperature anomalies. On the other hand, over the Sahara a significant surface warming seems associated with a slightly negative decrease in the sensible heat flux, a slightly negative surface eDRP (Figs. 3b and 4a), and a small positive anomaly in latent heat fluxes (Fig. 4d) corresponding to the region of increased precipitations (Fig. 2d).
During the NH winter the predominant effect of dust is a negative eDRP over the Sahel and the Atlantic Ocean (Fig. S8), associated to surface cooling and a high pressure anomaly over West Africa, with increasing surface divergence pushing the ITCZ farther south (Fig. S13).
The net TOA DRE and related temperature anomaly patterns (Figs. 1e,f and 2), characterized by an increased horizontal temperature gradient between land and ocean, lead to an enhanced summer monsoon. This spatial pattern, with a negative TOA DRE over the ocean and a positive TOA DRE over the desert, is common to other model studies using similar dust optical properties (Miller et al. 2004; Yoshioka et al. 2007; Balkanski et al. 2007), that is, relatively less absorbing, in accordance to a growing body of evidence from relatively recent observations (Kaufman et al. 2001; Sinyuk et al. 2003; McConnell et al. 2007; Müller et al. 2011; Moosmüller et al. 2012), as opposed to older, more absorbing optical properties that are nonetheless still in use in the community (Hess et al. 1998; Strong et al. 2015, and references therein). A realistic representation of both optical properties and particle size distributions is necessary for a realistic representation of dust–radiation interactions (Miller et al. 2006, 2014; Mahowald et al. 2014; Kok et al. 2017); observationally based estimates of dust clear-sky DRE efficiency (i.e., DRE per unit AOD) (Zhang and Christopher 2003; Li et al. 2004; Patadia et al. 2009; Song et al. 2018), ultimately attributable to a combination of both those features, are in agreement with our model results (Albani et al. 2014).
A few authors analyzed the sensitivity of dust impacts on surface climate as a function of dust optical properties (Tegen and Lacis 1996; Perlwitz et al. 2001; Miller et al. 2004), also in relation to the West African monsoon (Solmon et al. 2008; Zhao et al. 2011; Miller et al. 2014; Strong et al. 2015), confirming the different response of surface climate and the monsoon to different prescriptions in terms of dust optical properties. Miller et al. (2014) in particular also highlighted how vertical mixing and lateral energy transport differences among models further complicate the way in which dust DRE anomalies are translated into anomalies in surface climate variables.
We performed a sensitivity study using an outdated representation of dust in the CESM (Neale et al. 2010), still used by some in the community despite the results of Albani et al. (2014) showing the inconsistency with observational constraints (Albani et al. 2014; Sinyuk et al. 2003; Colarco et al. 2002). A large impact on atmospheric extinction is evident (Figs. S14 and S15), when comparing simulations using a finer-sized, more absorbing dust (old), as opposed to a less absorbing, coarser-sized dust (new) (Albani et al. 2014). This results in different patterns of TOA iDRE (SW only) and in a generalized surface cooling in the case of the finer-sized, more absorbing dust, and in the absence of dust–LW interactions (Fig. S16). A generalized positive anomaly in sea level pressure over North Africa and negative anomaly over the Gulf of Guinea drives a reduction of the intensity of the monsoon and strengthening of the African easterly jet (Figs. S17 and S18).
Because of the dependence of DRE on the surface albedo (Yu et al. 2006), past changes in vegetation, for instance during the mid-Holocene (Braconnot et al. 1999; Egerer et al. 2016), could modulate the dust DRE response (Pausata et al. 2016). In our case, prescribing a northward expansion of vegetation in the MH would likely result in dampening the effect of monsoon amplification by dust, which is already feeble given the much reduced dust load at this time.
Adopting very different strategies in terms of intrinsic optical properties and particle size distributions contributes to a very wide spread of model results (Huneeus et al. 2011; Boucher et al. 2013; Albani et al. 2018). We speculate that important uncertainties still exist in these parameters, as well as in the representation of clouds and vertical mixing in models, so that even similar TOA DRE fields may result in dissimilar impacts on surface temperature or precipitation (Miller et al. 2014; Strong et al. 2015) and variations in shape (Kalashnikova and Sokolik 2004; Potenza et al. 2016) and composition (Perlwitz et al. 2015a; Scanza et al. 2015) may still provide important uncertainty. Yet, several first-order constraints are available to the community (Yu et al. 2006; Huneeus et al. 2011; Albani et al. 2014; Ridley et al. 2016; Kok et al. 2017) that, if accounted for, could contribute to reduce the uncertainties in representing dust–climate interactions.
e. Dust variability impacts on changing climates
Figures 7 and 8 summarize the impacts that dust variability had over past climatic transitions, as depicted in our simulations. We consider idealized climate transitions as the successive anomalies between our four equilibrium climate simulations: deglaciation (MH − LGM), termination of the African Humid Period (PI − MH), and industrialization (CUR − PI). Ongoing climate change is represented by the anomaly in surface temperature or precipitation for the NODUST cases (e.g., NODUSTMH − NODUSTLGM) (Figs. 7a–c,g–i). Dust-related impacts on ongoing climate change are calculated as “double differences” [e.g., (DUST − NODUST)MH − (DUST − NODUST)LGM] (Figs. 7d–f,j–l).
Maps illustrating dust impacts in past climate transitions. Difference in annual (ANN) surface temperature (K) for the NODUST case between (a) MH and LGM, (b) PI and MH, and (c) CUR and PI. Dust-related difference in annual (ANN) surface temperature (K) between (d) MH and LGM, (e) PI and MH, and (f) CUR and PI. Difference in annual (ANN) total precipitations (mm day−1) for the NODUST case between (g) MH and LGM, (h) PI and MH, and (i) CUR and PI. Dust-related difference in annual (ANN) total precipitation (mm day−1) between (j) MH and LGM, (k) PI and MH, and (l) CUR and PI. Only significant anomalies are plotted [i.e., where the two-tailed FDR test is significant (alpha = 0.20)]. Dust-related differences in surface variables are defined as “double differences,” [e.g., (DUST-NODUST)MH − (DUST-NODUST)LGM].
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
Maps illustrating dust impacts in past climate transitions. Difference in annual (ANN) surface temperature (K) for the NODUST case between (a) MH and LGM, (b) PI and MH, and (c) CUR and PI. Dust-related difference in annual (ANN) surface temperature (K) between (d) MH and LGM, (e) PI and MH, and (f) CUR and PI. Difference in annual (ANN) total precipitations (mm day−1) for the NODUST case between (g) MH and LGM, (h) PI and MH, and (i) CUR and PI. Dust-related difference in annual (ANN) total precipitation (mm day−1) between (j) MH and LGM, (k) PI and MH, and (l) CUR and PI. Only significant anomalies are plotted [i.e., where the two-tailed FDR test is significant (alpha = 0.20)]. Dust-related differences in surface variables are defined as “double differences,” [e.g., (DUST-NODUST)MH − (DUST-NODUST)LGM].
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
Maps illustrating dust impacts in past climate transitions. Difference in annual (ANN) surface temperature (K) for the NODUST case between (a) MH and LGM, (b) PI and MH, and (c) CUR and PI. Dust-related difference in annual (ANN) surface temperature (K) between (d) MH and LGM, (e) PI and MH, and (f) CUR and PI. Difference in annual (ANN) total precipitations (mm day−1) for the NODUST case between (g) MH and LGM, (h) PI and MH, and (i) CUR and PI. Dust-related difference in annual (ANN) total precipitation (mm day−1) between (j) MH and LGM, (k) PI and MH, and (l) CUR and PI. Only significant anomalies are plotted [i.e., where the two-tailed FDR test is significant (alpha = 0.20)]. Dust-related differences in surface variables are defined as “double differences,” [e.g., (DUST-NODUST)MH − (DUST-NODUST)LGM].
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
As in Fig. 3, but for anomalies related to dust variability, defined as (DUST − NODUST)MH − (DUST − NODUST)LGM, during the NH summer (JJA). (a) Dust TOA eDRP (colors) and surface temperature anomalies (black line contours; intervals at 0 and ± 1, 2, 4, 8, 16 K); the dotted pattern highlights temperature anomalies whose absolute values are at least half of the corresponding absolutes value of the ongoing climate change anomaly (NODUSTMH − NODUSTLGM). (b) Anomalies in geopotential height (colors) and horizontal winds (arrows) at 700 hPa); a reference wind vector is marked in green, and its velocity is reported in the top-left corner. The statistically significant anomalies of the different variables shown in this figure are visible in Fig. S21.
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
As in Fig. 3, but for anomalies related to dust variability, defined as (DUST − NODUST)MH − (DUST − NODUST)LGM, during the NH summer (JJA). (a) Dust TOA eDRP (colors) and surface temperature anomalies (black line contours; intervals at 0 and ± 1, 2, 4, 8, 16 K); the dotted pattern highlights temperature anomalies whose absolute values are at least half of the corresponding absolutes value of the ongoing climate change anomaly (NODUSTMH − NODUSTLGM). (b) Anomalies in geopotential height (colors) and horizontal winds (arrows) at 700 hPa); a reference wind vector is marked in green, and its velocity is reported in the top-left corner. The statistically significant anomalies of the different variables shown in this figure are visible in Fig. S21.
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
As in Fig. 3, but for anomalies related to dust variability, defined as (DUST − NODUST)MH − (DUST − NODUST)LGM, during the NH summer (JJA). (a) Dust TOA eDRP (colors) and surface temperature anomalies (black line contours; intervals at 0 and ± 1, 2, 4, 8, 16 K); the dotted pattern highlights temperature anomalies whose absolute values are at least half of the corresponding absolutes value of the ongoing climate change anomaly (NODUSTMH − NODUSTLGM). (b) Anomalies in geopotential height (colors) and horizontal winds (arrows) at 700 hPa); a reference wind vector is marked in green, and its velocity is reported in the top-left corner. The statistically significant anomalies of the different variables shown in this figure are visible in Fig. S21.
Citation: Journal of Climate 32, 22; 10.1175/JCLI-D-18-0742.1
Figure 7 shows the variability of dust impacts across climatic periods, both globally and for key regions identified in this study, such as the Arctic and North Africa (Sahara and Sahel). The relative importance of regional impacts, compared to the global averages, is highlighted both in terms of eDRP and impacts on surface climate variables. In particular we see the opposite eDRP and surface temperature anomalies between the Sahara and the Sahel, and the enhanced West African monsoonal precipitation by dust, visible at any time except the MH when the dust loads were at minimum in the region. The most notable impact of dust variability is the negative feedback on Arctic surface temperature for the deglaciation analog (Figs. 7d and 8a,b).
The transition from the LGM to the (early and) mid-Holocene was characterized by a general warming (Fig. 7a), and by an increase in the hydrological cycle (Fig. 7g), as well as a significant decrease in dustiness worldwide, which is more marked at high and midlatitudes (Fig. 1a). The reduction in NH dust loads, and the associated cooling, partially offsets the warming trend from other forcings over this time period (Figs. 7d, 8a). In particular a strong negative eDRP is associated with lower dust loads over the Russian Arctic (Fig. 8a), accompanied by negative anomalies in geopotential heights at 700 hPa (Fig. 8b), which is consistent with the dust impacts on Arctic summers during the LGM, shown in Fig. 5. Negative anomalies (from dust) in surface temperature of several kelvin, particularly marked over western Siberia, are locally comparable in magnitude to the ongoing warming in response to changes in orbital parameters, greenhouse gases, sea level, and ice sheet distribution (Fig. 8a).
Dust-induced anomalies may have played a role in modulating the deglaciation patterns, which from the Greenland records we know were characterized by fast, almost synchronous variations of dust and temperatures (Mayewski et al. 1997). Because in Greenland colder conditions are associated with more dust (which there effectively acts as a warming agent), dust could have a negative feedback over the Arctic surface temperatures. The same mechanism, with opposite sign of the variables involved, may also be seen through a mirror when considering transitions from relatively warm, less dusty periods to colder stages with more dust mitigating the ongoing cooling over parts of the Arctic. This negative feedback may have been important to explain the abrupt transitions in dust and water oxygen isotopes between colder and warmer stages in Greenland ice cores (Steffensen et al. 2008). Interestingly, the dust-related surface temperature anomalies in our simulations are not seen in Greenland (Fig. 8a). Dust impacts on abrupt transitions may have taken the form of impacts on atmospheric circulation (Fig. 8b), in turn potentially impacting dust transport patterns (Mayewski et al. 1997) and dust emissions themselves (Ohgaito et al. 2018). Future studies, including the potential role of indirect aerosol–cloud impacts of dust on the LGM climate (Takemura et al. 2009; Sagoo and Storelvmo 2017), will need to clarify these aspects.
Other studies indicated either a positive dust TOA DRE in the LGM–PI comparison over the Arctic in general or Siberia (Overpeck et al. 1996; Claquin et al. 2003; Takemura et al. 2009), consistent with this work, or a negative DRE (Lambert et al. 2013; Hopcroft et al. 2015). As already discussed, these differences are largely attributable to differences in prescribed surface albedo, dust intrinsic optical properties, and particle size distributions, and modulated by the amount and spatial distributions of dust.
The dramatic increase in mid- and high-latitude dust sources during the glacial periods may have been linked to the development of important glaciogenic sources of dust (i.e., derived from glacier erosion) (Mahowald et al. 2006), and potentially also to dust source area expansion from CO2 suppression of vegetation primary production (Mahowald et al. 1999) and increased wind gustiness (McGee et al. 2010). While glaciogenic sources are very limited in extent and contribute very little to the global dust budget in warm climates such as today’s, ice melting at the edges of Greenland ice sheet may promote dust erosion (Bullard 2013). Considering the effective warming potential seen here (e.g., Figs. 4 and 5), it is possible that a positive feedback on ice melting could be activated locally at the margins of melting glaciers.
The PI–MH difference (an analog for the termination of the AHP) was characterized by a significant aridification and an associated increase in dust emissions from North Africa (Fig. 1a, Fig. S1), which in our simulations results in an annually averaged warming in Europe, although summers are cooler (Figs. 7b,e; see also Figs. S19 and S20).
Varying dust emissions since the LGM do not seem to have had statistically significant impacts on surface climate. However, note that changes in the hydrological cycle at low and middle latitudes between different climates, irrespective of dust, are not detected as significant in our simulations, according to the FDR test (Figs. 7h,i). Note that changes in vegetation not represented here may have had impacts on the physical climate (Claussen et al. 1999), and on dust direct feedbacks on climate through changes in surface albedo.
f. Long-term impacts of dust in glacial conditions
To estimate the importance of ocean coupling on the impact of dust on climate, we extended the LGM simulations (DUST and NODUST) for 150 additional years. There is a negative trend in global average surface temperatures, in response to an approximately constant negative TOA net radiative flux of −0.5 W m−2 (Fig. S22). This is the result of the initial conditions of our simulation (PMIP3 experiment) not being fully in climate equilibrium, as described in Brady et al. (2013). Still, the situation can be considered as a loose analog for a glaciation onset phase, allowing us to further explore dust–snow feedbacks.
In spite of the negative global surface temperature trends in both DUST and NODUST cases, we see that in the former case (DUST) the trend is slower, and is associated with a positive dust-induced anomaly (Fig. S22). The spatial features show that dust further accentuates the warming in the Siberian Arctic, along with more reduction of snow cover and increase in the leaf area index (Fig. S23). During the 200 years of the simulation, dust load increases by only 3% globally and it is unlikely to be the main driver of further effects, which instead we interpret as due to systems adjustments to dust forcing.
4. Summary
Despite important uncertainties from composition, size, and shape of the dust aerosols (Kalashnikova and Sokolik 2004; Kinne et al. 2006; Huneeus et al. 2011; Formenti et al. 2011; Mahowald et al. 2014; Scanza et al. 2015; Potenza et al. 2016), we do have first-order constraints that can be applied to many key variables in the representation of dust (Perlwitz et al. 2015b; Albani et al. 2014; Kok et al. 2017; Song et al. 2018), allowing for cross checks on resulting properties of the dust cycle and its impacts on climate. In previous studies, we deduced the best representation in our model of intrinsic properties of dust in CUR (Albani et al. 2014), and of the mass budgets of dust in four different climates (Albani et al. 2014, 2015, 2016). This allowed us to focus our attention on the direct effects of dust on surface climate here.
Our simulations show that, irrespective of the climate, the presence of dust has a tendency to cause a widespread cooling over the tropical oceans, and a warming over the major deserts and the Arctic. The potential of Arctic warming by dust is an intrinsic feature that derives from the optical properties of dust. This becomes manifest in the LGM, when both conditions are met of (i) relatively large dust loads at high NH latitudes and (ii) a bright surface in response to the presence of ice sheets and/or snow proximal to a massive dust source. In the tropics, the presence of dust results in an enhanced West African monsoon. Including dust direct effects seems therefore necessary for a realistic representation of monsoon dynamics. As a response of changing dust levels over the last 21 ka, we found that dust may have contributed to modulate climate evolution patterns in the Arctic, via a negative feedback.
We speculate that more observational constraints on dust properties and dust climate feedbacks will become available in the future, and that those will be extensively used to constrain Earth system models. This approach would contribute to reduce the uncertainty in estimating dust direct effects on climate, especially if accompanied by the use of reliable dynamical vegetation and ice sheet models, for two reasons. First, such a configuration would enable a time-evolving constraint on surface albedo, which modulates dust direct feedbacks on climate. Second, as shown in our analysis, dust impacts on the cryosphere and vegetation cover can be very important, at least during the LGM.
Acknowledgments
S.A. and N.M.M. acknowledge funding from NSF Grants 661541, 0932946, 1003509, and DOE-SC00006735. Computing resources were provided by the Climate Simulation Laboratory at NCAR’s Computational and Information Systems Laboratory (CISL), sponsored by the National Science Foundation and other agencies. We thank two anonymous reviewers for their constructive comments.
REFERENCES
Albani, S., and Coauthors, 2014: Improved dust representation in the Community Atmosphere Model. J. Adv. Model. Earth Syst., 6, 541–570, https://doi.org/10.1002/2013MS000279.
Albani, S., and Coauthors, 2015: Twelve thousand years of dust: The Holocene global dust cycle constrained by natural archives. Climate Past, 11, 869–903, https://doi.org/10.5194/cp-11-869-2015.
Albani, S., and Coauthors, 2016: Paleodust variability since the Last Glacial Maximum and implications for iron inputs to the ocean. Geophys. Res. Lett., 43, 3944–3954, https://doi.org/10.1002/2016GL067911.
Albani, S., Y. Balkanski, N. Mahowald, G. Winckler, V. Maggi, and B. Delmonte, 2018: Aerosol–climate interactions during the Last Glacial Maximum. Curr. Climate Change Rep., 4, 99–114, https://doi.org/10.1007/s40641-018-0100-7.
Atkinson, J. D., and Coauthors, 2013: The importance of feldspar for ice nucleation by mineral dust in mixed-phase clouds. Nature, 498, 355–358, https://doi.org/10.1038/nature12278.
Balkanski, Y., M. Schulz, T. Claquin, and S. Guibert, 2007: Reevaluation of mineral aerosol radiative forcings suggests a better agreement with satellite and AERONET data. Atmos. Chem. Phys., 7, 81–95, https://doi.org/10.5194/acp-7-81-2007.
Bartlein, P. J., and Coauthors, 2011: Pollen-based continental climate reconstructions at 6 and 21 ka: A global synthesis. Climate Dyn., 37, 775–802, https://doi.org/10.1007/s00382-010-0904-1.
Boucher, O., D. Randall, P. Artaxo, C. Bretherton, G. Feingold, and P. Forster, 2013: Clouds and aerosols. Climate Change 2013: The Physical Science Basis, T. F. Stocker et al., Eds., Cambridge University Press, 571–657.
Braconnot, P., S. Joussaume, O. Marti, and N. de Noblet, 1999: Synergistic feedbacks from ocean and vegetation on the African monsoon response to Mid-Holocene insolation. Geophys. Res. Lett., 26, 2481–2484, https://doi.org/10.1029/1999GL006047.
Brady, E. C., B. L. Otto-Bliesner, J. E. Kay, and N. Rosenbloom, 2013: Sensitivity to glacial forcing in the CCSM4. J. Climate, 26, 1901–1925, https://doi.org/10.1175/JCLI-D-11-00416.1.
Bullard, J. E., 2013: Contemporary glaciogenic inputs to the dust cycle. Earth Surf. Processes Landforms, 38, 71–89, https://doi.org/10.1002/esp.3315.
Claquin, T., and Coauthors, 2003: Radiative forcing of climate by ice-age atmospheric dust. Climate Dyn., 20, 193–202, https://doi.org/10.1007/s00382-002-0269-1.
Claussen, M., C. Kubatzki, V. Brovkin, A. Ganopolski, P. Hoelzmann, and H.-J. Pachur, 1999: Simulation of an abrupt change in Saharan vegetation in the Mid-Holocene. Geophys. Res. Lett., 26, 2037–2040, https://doi.org/10.1029/1999GL900494.
Colarco, P. R., O. B. Toon, O. Torres, and P. J. Rasch, 2002: Determining the UV imaginary index of refraction of Saharan dust particles from Total Ozone Mapping Spectrometer data using a three-dimensional model of dust transport. J. Geophys. Res., 107, 4289, https://doi.org/10.1029/2001JD000903.
Collins, W. J., and Coauthors, 2017: AerChemMIP: Quantifying the effects of chemistry and aerosols in CMIP6. Geosci. Model Dev., 10, 585–607, https://doi.org/10.5194/gmd-10-585-2017.
deMenocal, P., J. Ortiz, T. Guilderson, J. Adkins, M. Sarnthein, L. Baker, and M. Yarusinsky, 2000: Abrupt onset and termination of the African Humid Period. Quat. Sci. Rev., 19, 347–361, https://doi.org/10.1016/S0277-3791(99)00081-5.
Egerer, S., M. Claussen, C. Reick, and T. Stanelle, 2016: The link between marine sediment records and changes in Holocene Saharan landscape: Simulating the dust cycle. Climate Past, 12, 1009–1027, https://doi.org/10.5194/cp-12-1009-2016.
Eyring, V., S. Bony, G. A. Meehl, C. A. Senior, B. Stevens, R. J. Stouffer, and K. E. Taylor, 2016: Overview of the Coupled Model Intercomparison Project phase 6 (CMIP6) experimental design and organization. Geosci. Model Dev., 9, 1937–1958, https://doi.org/10.5194/gmd-9-1937-2016.
Flanner, M. G., C. S. Zender, J. T. Randerson, and P. J. Rasch, 2007: Present-day climate forcing and response from black carbon in snow. J. Geophys. Res., 112, D11202, https://doi.org/10.1029/2006JD008003.
Formenti, P., and Coauthors, 2011: Recent progress in understanding physical and chemical properties of African and Asian mineral dust. Atmos. Chem. Phys., 11, 8231–8256, https://doi.org/10.5194/acp-11-8231-2011.
Gent, P. R., and Coauthors, 2011: The Community Climate System Model version 4. J. Climate, 24, 4973–4991, https://doi.org/10.1175/2011JCLI4083.1.
Ginoux, P., M. Chin, I. Tegen, J. M. Prospero, B. Holben, O. Dubovik, and S.-J. Lin, 2001: Sources and distributions of dust aerosols simulated with the GOCART model. J. Geophys. Res., 106, 20 255–20 273, https://doi.org/10.1029/2000JD000053.
Ginoux, P., J. M. Prospero, T. E. Gill, N. C. Hsu, and M. Zhao, 2012: Global-scale attribution of anthropogenic and natural dust sources and their emission rates based on MODIS Deep Blue aerosol products. Rev. Geophys., 50, RG3005, https://doi.org/10.1029/2012RG000388.
Grodsky, S. A., J. A. Carton, and N. Sumant, 2003: Near surface westerly wind jet in the Atlantic ITCZ. Geophys. Res. Lett., 30, 2009, https://doi.org/10.1029/2003GL017867.
Hess, M., P. Koepke, and I. Schult, 1998: Optical properties of aerosols and clouds: The software package OPAC. Bull. Amer. Meteor. Soc., 79, 831–844, https://doi.org/10.1175/1520-0477(1998)079<0831:OPOAAC>2.0.CO;2.
Hopcroft, P. O., P. J. Valdes, S. Woodward, and M. M. Joshi, 2015: Last Glacial Maximum radiative forcing from mineral dust aerosols in an Earth system model. J. Geophys. Res., 120, 8186–8205, https://doi.org/10.1002/2015JD023742.
Huneeus, N., and Coauthors, 2011: Global dust model intercomparison in AeroCom phase I. Atmos. Chem. Phys., 11, 7781–7816, https://doi.org/10.5194/acp-11-7781-2011.
Jickells, T. D., and Coauthors, 2005: Global iron connections between desert dust, ocean biogeochemistry, and climate. Science, 308, 67–71, https://doi.org/10.1126/science.1105959.
Kageyama, M., and Coauthors, 2017: The PMIP4 contribution to CMIP6—Part 4: Scientific objectives and experimental design of the PMIP4-CMIP6 Last Glacial Maximum experiments and PMIP4 sensitivity experiments. Geosci. Model Dev., 10, 4035–4055, https://doi.org/10.5194/gmd-10-4035-2017.
Kageyama, M., and Coauthors, 2018: The PMIP4 contribution to CMIP6—Part 1: Overview and over-arching analysis plan. Geosci. Model Dev., 11, 1033–1057, https://doi.org/10.5194/gmd-11-1033-2018.
Kalashnikova, O. V., and I. N. Sokolik, 2004: Modeling the radiative properties of nonspherical soil-derived mineral aerosols. J. Quant. Spectrosc. Radiat. Transf., 87, 137–166, https://doi.org/10.1016/j.jqsrt.2003.12.026.
Kaplan, J. O., and Coauthors, 2003: Climate change and Arctic ecosystems: 2. Modeling, paleodata–model comparisons, and future projections. J. Geophys. Res., 108, 8171, https://doi.org/10.1029/2002JD002559.
Kaufman, Y. J., D. Tanré, O. Dubovik, A. Karnieli, and L. A. Remer, 2001: Absorption of sunlight by dust as inferred from satellite and ground-based remote sensing. Geophys. Res. Lett., 28, 1479–1482, https://doi.org/10.1029/2000GL012647.
Kinne, S., and Coauthors, 2006: An AeroCom initial assessment—Optical properties in aerosol component modules of global models. Atmos. Chem. Phys., 6, 1815–1834, https://doi.org/10.5194/acp-6-1815-2006.
Kohfeld, K. E., and S. P. Harrison, 2001: DIRTMAP: The geological record of dust. Earth Sci. Rev., 54, 81–114, https://doi.org/10.1016/S0012-8252(01)00042-3.
Kok, J. F., 2011: A scaling theory for the size distribution of emitted dust aerosols suggests climate models underestimate the size of the global dust cycle. Proc. Natl. Acad. Sci. USA, 108, 1016–1021, https://doi.org/10.1073/pnas.1014798108.
Kok, J. F., and Coauthors, 2017: Smaller desert dust cooling effect estimated from analysis of dust size and abundance. Nat. Geosci., 10, 274–278, https://doi.org/10.1038/ngeo2912.
Krinner, G., O. Boucher, and Y. Balkanski, 2006: Ice-free glacial northern Asia due to dust deposition on snow. Climate Dyn., 27, 613–625, https://doi.org/10.1007/s00382-006-0159-z.
Lambert, F., and Coauthors, 2008: Dust–climate couplings over the past 800,000 years from the EPICA Dome C ice core. Nature, 452, 616–619, https://doi.org/10.1038/nature06763.
Lambert, F., and Coauthors, 2013: The role of mineral-dust aerosols in polar temperature amplification. Nat. Climate Change, 3, 487–491, https://doi.org/10.1038/nclimate1785.
Lau, K. M., K. M. Kim, Y. C. Sud, and G. K. Walker, 2009: A GCM study of the response of the atmospheric water cycle of West Africa and the Atlantic to Saharan dust radiative forcing. Ann. Geophys., 27, 4023–4037, https://doi.org/10.5194/angeo-27-4023-2009.
Lavaysse, C., C. Flamant, S. Janicot, D. J. Parker, J.-P. Lafore, B. Sultan, and J. Pelon, 2009: Seasonal evolution of the West African heat low: A climatological perspective. Climate Dyn., 33, 313–330, https://doi.org/10.1007/s00382-009-0553-4.
Lawrence, D. M., and Coauthors, 2011: Parameterization improvements and functional and structural advances in version 4 of the Community Land Model. J. Adv. Model. Earth Syst., 3, M03001, https://doi.org/10.1029/2011MS00045.
Li, F., A. M. Vogelmann, and V. Ramanathan, 2004: Saharan dust aerosol radiative forcing measured from space. J. Climate, 17, 2558–2571, https://doi.org/10.1175/1520-0442(2004)017<2558:SDARFM>2.0.CO;2.
Maher, B. A., J. M. Prospero, D. Mackie, D. Gaiero, P. P. Hesse, and Y. Balkanski, 2010: Global connections between aeolian dust, climate and ocean biogeochemistry at the present day and at the Last Glacial Maximum. Earth Sci. Rev., 99, 61–97, https://doi.org/10.1016/j.earscirev.2009.12.001.
Mahowald, N. M., K. Kohfeld, M. Hansson, Y. Balkanski, S. P. Harrison, I. C. Prentice, M. Schulz, and H. Rodhe, 1999: Dust sources and deposition during the last glacial maximum and current climate: A comparison of model results with paleodata from ice cores and marine sediments. J. Geophys. Res., 104, 15 895–15 916, https://doi.org/10.1029/1999JD900084.
Mahowald, N. M., D. R. Muhs, S. Levis, P. J. Rasch, M. Yoshioka, C. S. Zender, and C. Luo, 2006: Change in atmospheric mineral aerosols in response to climate: Last glacial period, preindustrial, modern, and doubled carbon dioxide climates. J. Geophys. Res., 111, D10202, https://doi.org/10.1029/2005JD006653.
Mahowald, N. M., S. Albani, J. F. Kok, S. Engelstaeder, R. Scanza, D. S. Ward, and M. G. Flanner, 2014: The size distribution of desert dust aerosols and its impact on the Earth system. Aeolian Res., 15, 53–71, https://doi.org/10.1016/j.aeolia.2013.09.002.
Mahowald, N. M., R. Scanza, J. Brahney, C. L. Goodale, P. G. Hess, J. K. Moore, and J. Neff, 2017: Aerosol deposition impacts on land and ocean carbon cycles. Curr. Clim. Change Rep., 3, 16–31, https://doi.org/10.1007/s40641-017-0056-z.
Mayewski, P. A., L. D. Meeker, M. S. Twickler, S. Whitlow, Q. Yang, W. B. Lyons, and M. Prentice, 1997: Major features and forcing of high-latitude Northern Hemisphere atmospheric circulation using a 110,000-year-long glaciochemical series. J. Geophys. Res. Oceans, 102, 26 345–26 366, https://doi.org/10.1029/96JC03365.
McConnell, J. R., and Coauthors, 2007: 20th-century industrial black carbon emissions altered Arctic climate forcing. Science, 317, 1381–1384, https://doi.org/10.1126/science.1144856.
McGee, D., W. S. Broecker, and G. Winckler, 2010: Gustiness: The driver of glacial dustiness? Quat. Sci. Rev., 29, 2340–2350, https://doi.org/10.1016/j.quascirev.2010.06.009.
McGee, D., P. B. deMenocal, G. Winckler, J. B. W. Stuut, and L. I. Bradtmiller, 2013: The magnitude, timing and abruptness of changes in North African dust deposition over the last 20,000 yr. Earth Planet. Sci. Lett., 371–372, 163–176, https://doi.org/10.1016/j.epsl.2013.03.054.
Miller, R. L., and I. Tegen, 1998: Climate response to soil dust aerosols. J. Climate, 11, 3247–3267, https://doi.org/10.1175/1520-0442(1998)011<3247:CRTSDA>2.0.CO;2.
Miller, R. L., J. P. Perlwitz, and I. Tegen, 2004: Feedback upon dust emission by dust radiative forcing through the planetary boundary layer. J. Geophys. Res., 109, D24209, https://doi.org/10.1029/2004JD004912.
Miller, R. L., and Coauthors, 2006: Mineral dust aerosols in the NASA Goddard Institute for Space Sciences ModelE atmospheric general circulation model. J. Geophys. Res., 111, D06208, https://doi.org/10.1029/2005JD005796.
Miller, R. L., P. Knippertz, C. Pérez García-Pando, J. P. Perlwitz, and I. Tegen, 2014: Impact of dust radiative forcing upon climate. Mineral Dust, P. Knippertz and J.-B.W. Stuut, Eds., Springer, 327–357.
Moosmüller, H., J. P. Engelbrecht, M. Skiba, G. Frey, R. K. Chakrabarty, and W. P. Arnott, 2012: Single scattering albedo of fine mineral dust aerosols controlled by iron concentration. J. Geophys. Res., 117, D11210, https://doi.org/10.1029/2011JD016909.
Müller, T., A. Schladitz, K. Kandler, and A. Wiedensohler, 2011: Spectral particle absorption coefficients, single scattering albedos and imaginary parts of refractive indices from ground based in situ measurements at Cape Verde Island during SAMUM-2. Tellus, 63B, 573–588, https://doi.org/10.1111/j.1600-0889.2011.00572.x.
Neale, R. B., J. H. Richter, A. J. Conley, S. Park, P. H. Lauritzen, and A. Gettleman, 2010: Description of the NCAR Community Atmosphere Model (CAM 4.0). NCAR Tech. Note NCAR/TN-485+STR, 212 pp., www.cesm.ucar.edu/models/ccsm4.0/cam/docs/description/cam4_desc.pdf.
Ohgaito, R., A. Abe-Ouchi, R. O’ishi, T. Takemura, A. Ito, T. Hajima, S. Watanabe, and M. Kawamiya, 2018: The effect of high dust amount on the surface temperature during the Last Glacial Maximum: A modelling study using MIROC-ESM. Climate Past, 14, 1565–1581, https://doi.org/10.5194/cp-14-1565-2018.
Otto-Bliesner, B. L., and Coauthors, 2017: The PMIP4 contribution to CMIP6—Part 2: Two interglacials, scientific objective and experimental design for Holocene and Last Interglacial simulations. Geosci. Model Dev., 10, 3979–4003, https://doi.org/10.5194/gmd-10-3979-2017.
Overpeck, J., D. Rind, A. Lacis, and R. Healy, 1996: Possible role of dust-induced regional warming in abrupt climate change during the last glacial period. Nature, 384, 447–449, https://doi.org/10.1038/384447a0.
Patadia, F., E.-S. Yang, and S. A. Christopher, 2009: Does dust change the clear sky top of atmosphere shortwave flux over high surface reflectance regions? Geophys. Res. Lett., 36, L15825, https://doi.org/10.1029/2009GL039092.
Patricola, C. M., and K. H. Cook, 2007: Dynamics of the West African monsoon under mid-Holocene precessional forcing: Regional climate model simulations. J. Climate, 20, 694–716, https://doi.org/10.1175/JCLI4013.1.
Pausata, F. S. R., G. Messori, and Q. Zhang, 2016: Impacts of dust reduction on the northward expansion of the African monsoon during the Green Sahara period. Earth Planet. Sci. Lett., 434, 298–307, https://doi.org/10.1016/j.epsl.2015.11.049.
Pérez García-Pando, C., and Coauthors, 2014: Soil dust aerosols and wind as predictors of seasonal meningitis incidence in Niger. Environ. Health Perspect., 122, 679–686, https://doi.org/10.1289/EHP.1306640.
Perlwitz, J., I. Tegen, and R. L. Miller, 2001: Interactive soil dust aerosol model in the GISS GCM: 1. Sensitivity of the soil dust cycle to radiative properties of soil dust aerosols. J. Geophys. Res., 106, 18 167–18 192, https://doi.org/10.1029/2000JD900668.
Perlwitz, J., C. Pérez García-Pando, and R. L. Miller, 2015a: Predicting the mineral composition of dust aerosols—Part 1: Representing key processes. Atmos. Chem. Phys., 15, 11 593–11 627, https://doi.org/10.5194/acp-15-11593-2015.
Perlwitz, J., C. Pérez García-Pando, and R. L. Miller, 2015b: Predicting the mineral composition of dust aerosols—Part 2: Model evaluation and identification of key processes with observations. Atmos. Chem. Phys., 15, 11 629–11 652, https://doi.org/10.5194/acp-15-11629-2015.
Potenza, M. A. C., and Coauthors, 2016: Shape and size constraints on dust optical properties from the Dome C ice core, Antarctica. Sci. Rep., 6, 28162, https://doi.org/10.1038/srep28162.
Prospero, J. M., P. Ginoux, O. Torres, S. E. Nicholson, and T. E. Gill, 2002: Environmental characterization of global sources of atmospheric soil dust identified with the NIMBUS 7 Total Ozone Mapping Spectrometer (TOMS) absorbing aerosol product. Rev. Geophys., 40, 1002, https://doi.org/10.1029/2000RG000095.
Pu, B., and K. H. Cook, 2010: Dynamics of the West African westerly jet. J. Climate, 23, 6263–6276, https://doi.org/10.1175/2010JCLI3648.1.
Pu, B., and K. H. Cook, 2012: