Carbon dioxide removal through enhanced weathering of basalt on agricultural land –Assessing the potential in Austria

Enhanced weathering through basalt application on agricultural land represents a proposed strategy for the removal of carbon dioxide from the atmosphere. It has been shown that enhanced weathering is principally feasible on a global scale, but it remains unclear whether it can be implemented on a local level. This information is however vital, to evaluate, if enhanced weathering should be further considered as a factor to alleviate the impact of the climate crisis. With this in mind, this article reviews of the current state of research and estimates the potential for CO2 removal on regional scale through a case study for Austria. Scenarios are estimated for three different particle size distributions (< 100 μm, < 10 μm and < 1 μm). Transport related emissions may cancel out any drawdown if grain sizes (< 100 μm) are used. However, under optimal transport conditions the largescale application of particles with a diameter < 10 μm may remove about 2% of Austria's annual Greenhouse gas emissions. We discuss challenges towards this goal, including the enormous amounts of basalt needed and the energy requirement related to grinding, as well as uncertainties related to actual field weathering rates. Those uncertainties hinder the precise quantification of CO2 drawdown as of now. While enhanced weathering remains a promising path for climate change mitigation, further research at laboratory and field scale is required to put this technology to optimal use.


Introduction
Within the framework of the Paris Agreement, the signing nations have agreed to keep global warming below 2 °C with respect to the pre-industrial level and to pursue efforts to limit the increase to 1.5 °C above pre-industrial levels to mitigate the effects of human caused climate change.
Estimates of the remaining global carbon budget consistent with a 1.5 °C and 2 °C scenario imply that the annual global output of total emissions should be curbed down to 25 GtCO2e (carbon dioxide equivalent) and 41 GtCO2e by the year 2030, respectively (Rogelj et al. 2018). However, within the framework of current policies, the estimated annual output in 2030 will be around 60 GtCO2e and even if countries manage to reach their self-determined contributions to reduce emissions, this value will only lower to about 54 GtCO2e (Emissions Gap Report 2019. With the remaining carbon budget rapidly diminishing, negative emission technologies (NET) are increasingly coming into focus.
Instead of reducing CO2 production, NETs focus on removing CO2 from the atmosphere and storing it in oceanic and terrestrial sinks. Current approaches include bioenergy with carbon capture and storage (BECCS), direct air capture (DAC), Enhanced Rock Weathering (EW), afforestation, ocean fertilization and the conversion of biomass into biochar for soil amendment (Negative Emissions Technologies and Reliable Sequestration: A Research Agenda, 2019;Smith et al., 2016). Large scale deployment of negative emission technologies within the 21 st century is an integral part of the vast majority of models, consistent with the 1.5 °C and 2 °C scenario (Workman et al., 2020). These models often imply large scale removal of atmospheric CO2 by NETs only after 2050, allowing a temporal overshoot of our remaining carbon budget (Hilaire et al., 2019). It has been argued that within the framework of the current policies it might be impossible to reach the 1.5 °C goal without relying on negative emission technologies (Anderson et al., 2020;Hickel and Kallis, 2020). Until now different NETs are at different stages of development (Anderson and Peters, 2016) and at the moment there is no single solution for the removal of CO2 at scale ready for deployment (Fuss et al., weathering (EW), or Terrestrial Enhanced Weathering (TEW). We use the term EW in this study. The concept has since been investigated in a number of different scenarios. The potential of mine tailings from mafic and ultramafic deposits has been evaluated by (e.g. Assima et al., 2014Assima et al., , 2012Gras et al., 2020;Hamilton et al., 2018;Harrison et al., 2013;Lechat et al., 2016;Oskierski et al., 2016;Power et al., 2020;Pronost et al., 2011;Thom et al., 2013;Wilson et al., 2014). Investigations include numerical modelling studies on global (Hartmann et al., 2013;Köhler et al., 2010;Moosdorf et al., 2014;Strefler et al., 2018;Taylor et al., 2016) and regional scale (Lefebvre et al., 2019;Renforth, 2012). While some studies focus on the application to croplands (Beerling, 2017;Haque et al., 2019;Kantola et al., 2017;Taylor et al., 2017) others focus on the marine environment and the effect of EW on ocean alkalinity (Bach et al., 2019;Montserrat et al., 2017;Renforth and Henderson, 2017;Rigopoulos et al., 2018a). Additonally, several field and lab studies have been carried out so far (Amann et al., 2020;Dietzen et al., 2018;Haque et al., 2020aHaque et al., , 2020bKelland et al., 2020;Renforth et al., 2015;ten Berge et al., 2012). One advantage of EW is relatively low land use compared to other NETs and low use of the resource water (Smith et al., 2016). Moreover, enhanced weathering of dunite and basalt is cost competitive with respect to other methods, ranging from 60 US $ per t CO2 for dunite to 200 US $ for basalt (Strefler et al., 2018). While the relatively high energy input required for EW related to mining, milling and transport of the rocks reduces efficiency, the CO2 drawdown of olivine may still reach a net value of 0.5 -1 t CO2 per t of rock (Beerling et al., 2018;Moosdorf et al., 2014). Olivine is used in many studies related to EW for its high CO2 drawdown relative to rock mass (e.g. Dietzen et al., 2018;Renforth et al., 2015;table 1), but the potential release of toxic metals during dissolution is a disadvantage (Amann et al., 2020). In contrast, Haque and coworkers showed in a number of studies that wollastonite (CaSiO3) amendment of soil is beneficial for crop yield, while at the same increasing soil inorganic carbon content (Haque et al., 2020b(Haque et al., , 2020a(Haque et al., , 2019. Similarily, basalt is already used for soil amelioration and results from an experimental study on the use of basalt powder, added to slightly acidic clay-loam soil, show cumulative removal of 4 t CO2 ha -1 within 5 years (Kelland et al., 2020). This study also found an increasing crop yield (sorghum) by 21 ± 9.4 %. In a recent study Beerling et al. (2020) indicated the potential of basalt application to cropland for CO2 removal on a global scale. They estimated the potential of CO2 removal for application of ground basaltic rock on cropland in the range of about 0.5 -2 GtCO2 y -1 . The aggregate amount of carbon dioxide removal, if sustained over 50 years could be between 25 and 100 Gt CO2. The study also showed the potential for removal of CO2 in temperate climate regions. Factors influencing the potential of EW include climate, agricultural area and practice, availability of suitable rocks, available transport networks, the CO2 intensity of electricity generation, and CO2 emissions from mining, comminution and transport. The importance of transport distance has been addressed by Lefebvre et al. (2019). The authors assessed the potential of basalt for EW in a part of Brazil and concluded that the maximum transport between the quarry and field would 990 ± 116 km, above which the transport related CO2 emissions would offset the potential CO2 capture. Therefore, if EW is to be used in the near future, the regional availability of suitable resources is a prerequisite and regionally tailored approaches are needed to put the technology to optimal use. Here we selected Austria as suitable case example to evaluate the feasibility of EW in a regional context. Towards the realization of the European Green Deal, countries with a relatively high share of hydropower for electricity generation, like Austria, could be well suited for EW as a technology. However, it remains to be tested whether enough basaltic rocks are available in Austria or bordering countries and whether the energy demand for mining, transport and mineral comminution and consequent CO2 production substantiate enhanced basalt weathering as a reliable strategy. In the following, we will first review the current state of research in the field before focusing on our case study.

Co-benefits from basaltic soil amelioration
One of the goals proclaimed in the Paris agreement on climate change is to"… foster climate resilience and low greenhouse gas emissions development, in a manner that does not threaten food production". In this context, the potential co-benefits of soil amelioration through application of basaltic rocks to croplands are noteworthy. They largely contrast the societal friction points, related to competition for land use, connected to BECCS, afforestation and reforestation. In addition, EW could be co-deployed for the two latter technologies (Amann and Hartmann, 2019;Kantola et al., 2017). Application of silicate -rock fertilizers to agricultural land is not a new concept and it has been investigated already by e.g. (Hensel, 1894) or more recently by e.g. (Van Straaten, 2006). The potential benefits are linked to the fact that silicon is actively accumulated in a number of crops including wheat, barley and sugar beet (Guntzer et al., 2012). Benefits of Si are especially important under environmental stress because silicon alleviates toxicity of some toxic metals (Adrees et al., 2015), improves Potassium, Phosphor and Calcium uptake, alleviates the effects of drought and increases resistance to pathogens and insects. The recycling of plant silicon in unaltered ecosystems accounts for a high percentage in the silicon pool, owed to the re-dissolution of plant phytoliths.
Under agricultural use, however, this cycle is interrupted due to the removal of crops. Especially with silicon accumulating plants this can warrant silicon fertilization (Savant et al., 1997) and a positive effect on yield has been reported for instance in the case of wheat with a grain yield increase of 4.1 -9.3 % (Liang et al., 1994). Soil amendment through application of basalt has positive effects on the cation exchange capacity in soils and supplies Potassium, Phosphor and Calcium (Gillman et al., 2002). In addition Aluminum toxicity is a problem with acid soil conditions around the world. Silicon in general and basalt have been shown to effectively decrease Al toxicity in acid soils ( Cocker et al., 1998;Shamshuddin and Kapok, 2010). For these potential co-benefits, basaltic rock may be the best candidate for EW through application on agricultural land, despite its relatively low capability to sequester CO2, compared to dunite. Moreover release of Chromium, Nickel and Cobalt from dunite (Renforth et al., 2015;ten Berge et al., 2012) and serpentinites (Kierczak et al., 2016) and potential accumulation in soil and crops may counteract the intended soil amelioration and the same is true for low Ca/Mg ratios (Echevarria, 2018). While serpentinites and amphiboles might exhibit good reactivity, asbestos group minerals are generally a safety concern during mining, milling, and spreading of rock powder, thereby even if handled in a safe way, public perception could impose barriers on their application (Gwenzi, 2020).

CO2 drawdown through weathering
Over geological time scales, atmospheric carbon dioxide is affected by the weathering of silicate rocks through the transformation of CO2 into HCO3 - (Amiotte Suchet et al., 2003;Berner et al., 1983;Dessert et al., 2003;Gaillardet et al., 1999;Gislason et al., 2009;Walker et al., 1981) . For instance, the rise of the Himalayas and consequent increased weathering has resulted in CO2 drawdown and a general cooling of global climate (Raymo and Ruddiman, 1992). Similarly, past cooling phases of the climate since the Archean could be directly linked to the drawdown of CO2 from the atmosphere (e.g. Elliot Smith et al., 2008;Kump et al., 1999;Lowe and Tice, 2004). The idea behind EW is to speed up this process through grinding of rocks and targeted application to increase weathering rates.
Exemplarily the chemical reaction can be followed through incongruent dissolution of anorthite (eq. 1). The dissolution of primary silicates leads to formation of secondary precipitates, releasing cations and transforming CO2 into HCO3 -.
Carbonate formation is an important mechanism for the in situ fixation of CO2 through Carbon capture and storage (For a review of mineral carbonation in general see e.g. Kelemen et al., 2020;Snaebjörnsdóttir et al., 2020). However, the aim of enhanced weathering is to convert CO2 into alkalinity, as the formation of carbonates will reduce the processes efficiency (eq. 2).
The maximum amount of CO2 drawn from the atmosphere through silicate dissolution is a function of the cation flux (mostly Ca 2+ , Mg 2+ , K + and Na + ) which is charge balanced by the formation of HCO3 -.
Drawdown potential can be expressed as the amount of CO2 (RCO2) removed from the atmosphere per mass of rock (t CO2 t -1 of rock). This value might be reduced during riverine transport (Hotchkiss et al., 2015;Marx et al., 2017;Polsenaere et al., 2013) and it will further decrease, depending on the apparent carbonate equilibrium in the ocean . For the relatively fast weathering of Ca and Mg bearing silicates, RCO2 is sometimes based solely on Mg 2+ and Ca 2+ contents (Renforth, 2012). In the case of basalt the contribution of monovalent cations (Na + + K + ) needs to be included, as they are frequently present in basaltic rocks (see table 1) in the form of relatively fast weathering minerals such as nepheline (Tole et al., 1986). In addition, the calculation of RCO2 from incongruent dissolution, is based on the idea of Al conservation through the formation of secondary minerals (eq. 1). However, congruent dissolution of the aluminosilicate will precede the formation of the secondary phase (Maher et al., 2009;Steefel and Van Cappellen, 1990) (eq. 3) and far from equilibrium conditions can be sustained during basalt weathering for instance through complexation of Al 3+ with organic acids (Perez-Fodich and Derry, 2019).  Table 1 provides the RCO2 for minerals commonly found in basalt and two different compositions of basalt (Gudbrandsson et al., 2011;Navarre-Sitchler and Brantley, 2007). RCO2 low is based on incongruent dissolution, where CO2 is charge balanced against the cations released during dissolution (ΣCO2 = meq(Ca 2+ + Mg 2+ + Na + + K + )). RCO2 max is based on congruent dissolution including AL 3+ , Fe 3+ , Fe 2+ and Ti 4+ in the calculation. The latter process has also been proposed in (Navarre-Sitchler and Brantley, 2007) to asses CO2 drawdown from basalt weathering. The values of 282 kg t -1 of basalt (RCO2 low) and 880 kg t -1 of basalt (RCO2 max) will be used for evaluation of the potential in a pessimistic and optimistic scenario in this study.

Basalt weathering rates in the context on enhanced weathering
The dissolution of basaltic glass at far from equilibrium conditions has been investigated in mixed flow reactors by (Galeczka et al., 2014;Gislason and Oelkers, 2003;Oelkers and Gislason, 2001;Stockmann et al., 2011;Wolff-Boenisch et al., 2004). These studies show a minimum dissolution rate in the circumneutral pH area (see figure 1), commonly observed for aluminosilicates (Gautier et al., 1994;Gudbrandsson et al., 2014). In contrast, Al-free silicate minerals, such as enstatite or forsterite exhibit a near linear dependence of dissolution rates on pH (figure 1). Accordingly, the respective mineralogical composition of a basaltic rock may influence the stoichiometry of the released cations.
To understand better the importance of mafic and ultramafic rocks as reactants in the long term global Silicon Carbon cycle, basalt weathering rates have been estimated at catchment scale in several studies (e.g. Babechuk et al., 2014;Das et al., 2005;Dessert et al., 2001;Li et al., 2016;Louvat and Allègre, 1997). However, reconciling the observations from field scale with the laboratory derived weathering rates remains a challenge. In general the difference in laboratory and field rates can be attributed to both, physical i.e. hydrological features (Velbel, 1993), chemical features such as surface passivation though leached layer formation or the precipitation of secondary phases (Daval et al., 2018;Velbel, 2009), or a mixture of both, such as a long fluid residence times. The prediction of surface passivation itself is not straightforward. For instance the formation of passivating layers may be prevented through fungal biofilm, thereby increasing dissolution rates (Gerrits et al., 2020).
On the other hand, the passivating effect of Al on the solubility and dissolution rate of amorphous silica and quartz has been investigated by (Bickmore et al., 2006;Iler, 1973), showing that Al containing layers not only reduce dissolution rates but also apparent overall solubility. Fluid residence time, soil moisture and mineral/fluid ratios are crucial when investigating the rate of chemical weathering (Maher, 2010;Navarre-Sitchler et al., 2011;White and Brantley, 2003). The complex interrelationship between gas and water in the unsaturated zone further complicates the calculation of realistic weathering rates (Harrison et al., 2017(Harrison et al., , 2015. In addition, field weathering rates are influenced by trace metals (Oelkers et al., 2018), bacterial communities (Wild et al., 2019) or organic ligands (Perez-Fodich and Derry, 2019;Perez et al., 2015). They are not only dependent on a specific surface area but can be influenced by crystallographic orientation of minerals (Daval et al., 2013).
In the context of enhanced weathering, dissolution of rock powder added to soil cannot be calculated merely based on dissolution rates. The shrinking core model has been used to describe the time for complete dissolution of a grain ). The authors estimated that the addition of olivine to seawater requires particle sizes < 10 μm to achieve relatively complete dissolution of 100 years. With a similar approach Renforth et al. (2015) estimated that the particle size of olivine added to soil, would have to be in the range of 0.1 -0.01 μm to dissolve within 5 years.
Notwithstanding the uncertainties related to field weathering, in this study dissolution was calculated, using a dissolution rate from (Gudbrandsson et al., 2011), who measured a BET normalized rate of 3.55 10 -12 mol/m²/s at pH 5.84 and 25 °C (see figure 1). This is probably at the high end of weathering rates, given the discrepancy between lab and field rates and 25 °C is certainly far away from the annual average temperature in Austria. It is however close to the high end of dissolution rates reported for olivine in soil by (Renforth et al., 2015) (see figure 1) and it is also in the range of the elemental release rates reported for basalt added to soil (Kelland et al., 2020). These authors measured an elemental Mg release rate of 6.6 x 10 -13 mol m -2 s -1 and a Ca release of 2.8 x 10 -12 mol m -2 s -1 . For comparison, the stoichiometric dissolution of our model basalt (table 1)  The BET surface area of the model basalt was calculated using an empirical equation, provided in (Brantley and Mellott, 2000) for olivine, where d is the particle diameter. log Molecular weight of the basalt is 125 g/m² and the density is assumed 3g/cm³. To apply the shrinking core model, based on spherical particles the BET normalized ratio has to be recalculated for geometric surface. Toward this goal the geometric surface area (SSAgeo) of spherical grains of each fraction was calculated according to the equation provided in (Tester et al., 1994). The roughness ratio (SSABET/SSAGeo) was then used to recalculate the dissolution rates based on geometric surface area (see the supplementary file for details about the calculation).

Availability of volcanic rocks in Austria and neighboring countries
Potential occurrences of volcanic rocks in the vicinity of Austria comprise units from the Western and Central European volcanic province in the Massif Central (Lustrino and Wilson, 2007), the Bohemian Massif and Eger Graben (Ulrych et al., 2011), Germany (Jung et al., 2005;Jung and Masberg, 1998), the Pannonian-Carpathian Volcanic province -parts of which are based in SE Austria (Ali et al., 2013;Downes et al., 1995;Lukács et al., 2018;Seghedi et al., 2004) and the Veneto Volcanic province in Northern Italy (Peccerillo, 2005) as well as some occurrences in the Dinaric alps (Prelević et al., 2005), (see figure 2).   (Cortesogno et al., 1998). The same is true for relatively large volcanic fields close in the West Carpathians, such as the central Slovakian volcanic field, mostly comprising andesite (Chernyshev et al., 2013). Similarly, volcanic rocks of the roman volcanic province are largely of intermediate composition (Beccaluva et al., 1991). (Plata et al., 2021) investigated the potential of andesitic and dacitic mining waste as a soil fertilizer, showing that soil amelioration is not limited to volcanic rocks with a low Si content.

Agricultural areas in Austria
Cropland area covered approximately 1,325 M ha in 2019 (Statistik Austria, 2020), and additional 0,576 M ha were categorized as intensive grassland (Statistik Austria, 2018), assumed to be easily available for inorganic fertilizer application. Out of 1152 representative soil samples 38% have a soil pH >7, in 53 % of samples the pH ranges from 5-7 and 9% are characterized as strongly acidic (pH = 4-5). For grass land only 6 % show a pH > 7, 55% of the samples are in the range 5-7 and 39% are between pH 4 and pH 5 (Schwarz and Freudenschuss, 2004). Soil amelioration is mostly proposed for acidic soils.
However, from a perspective of CO2 sequestration, soil pH has relatively low impact as long as basalt is used for soil amelioration. The dissolution rate of basaltic glass in the relevant range (4-8.5) reaches a minimum around pH 6 (see figure 1). Compared to this minimum, it is ~3 times faster compared at pH 5 and it is also ~3 times faster at pH 8. A pronounced pH effect would only be expected for very acid conditions -at pH 4 the dissolution rate is ~17 times that of pH 6 ( Gislason and Oelkers, 2003).
From this perspective basaltic soil amelioration could also be applied to lime rich alkaline soils which contribute important agricultural areas in Austria, where it could provide a Potassium source, decreasing the need for conventional fertilizers. In summary, an area of 1,902 million ha (cropland plus intensive grassland) are assumed a reasonable maximum area of application and will be used in further modelling exercises.

Mining, transport and comminution
The energy demand related to crushing and grinding of material in order to reach particle sizes, small enough to achieve desirable weathering rates in soil are considered and important drawback of EW (Gerdemann et al., 2007;Moosdorf et al., 2014;Renforth et al., 2015;Rigopoulos et al., 2018b). To evaluate the interrelationship between CO2 emissions from comminution and drawdown rate of CO2, three scenarios with different grain size distribution are used for modelling in the range 100 -0,1 μm (< 100 μm), 10 -0,1 μm (< 10 μm) and 1-0,1 μm (< 1 μm). The particle size distribution was calculated for fixed classes, with normal distribution with a μ of 50, 5 and 0,5 and σ of 30,3 and 0.3, respectively. The resulting grain size distribution can be found in the supplementary file (table S1).
Comminution, mining and application can be considered static and resulting input parameters, used for calculation of the CO2 balance are listed in Table 2. Based on the relatively widespread availability of mafic and ultramafic volcanic rocks (see above) the average distance from mine to field was estimated roughly about 300 km. This is an optimistic approach, assuming that the basalt quarries closest to the Austrian border are also able to produce the required amount and quality. Resulting transport related CO2 emissions considered in this study, are 4.8 kg CO2e t -1 , 18.9 kg CO2e t -1 and 59.4 kg CO2e t -1 , for transport by rail, a < 7.5t truck and 20 -40t truck, respectively. Total CO2e emissions including mining, comminution, transport and application calculated on base of 300km transport are shown in table 3 for three scenarios, using different grain size distributions (< 100 μm, < 10 μm, < 1 μm).

Influence of particle size on weathering rates
Based on Eq. 4 the average SSABET resulting is 10.53 (< 100 μm), 18.61 (< 10 μm) and 67.42 m²/g (< 1 μm), respectively which is in good agreement with the SSABET of around 10.5 m 2 /g, measured by Kelland et al. (2020) in the < 90 μm fraction of their basalt powder. Time for complete dissolution of all basalt was calculated to be 1696, 134 and 10.7 years for the < 100 μm, < 10 μm and < 1 μm powder, respectively. As the reservoir gets depleted and the grains become increasingly small, dissolution becomes relatively ineffective and for instance 87% of a 100 μm grain will have dissolved in half the time, needed for complete dissolution. Accordingly, the evaluated timeframe in this study  is for 97.5% dissolution in which case the dissolution takes 1205, 96 and 7.6 years (Table S1). As this time is based on the largest size grains in each powder, smaller grains will dissolve completely and based on the chosen grain size distribution > 99.9 % of the basalt powder will dissolve within the chosen timeframe already (see figure S1 in the supplementary file). The potential CO2 drawdown associated with this dissolution is shown in figures 4a -4c for the first 10 years and RCO2 max and RCO2 low. Grey dashed lines denote the estimates for CO2 emissions, generated through mining, grinding, transport (300km) and application; the low estimate is based on railway transport; the high estimate is based on road transport using a <7.5t truck.
Based on their experiments, Kelland et al. (2020) modelled a cumulative removal of CO2 of 44 kg CO2 t -1 of basalt within 5 years through the application of basalt with a diameter of < 128 μm. This drawdown was not linear and the largest fraction was drawn down during the 120 days experimental run (30 kg CO2 t -1 basalt) in good agreement with our calculation for the < 100 μm (figure 4a).
Within the chosen observation period, increasing the reaction rate through decreasing the grain size is decisive in the RCO2 low and the RCO2max scenario (see figure 5). Apparently comminution, despite its high energy demand is not the controlling factor on the CO2 drawdown. On the contrary, given the slow dissolution of the < 100 μm fraction (1698 years for complete dissolution), even best practice (rail transport) will only contribute limited to climate change mitigation (see table S2 and figure 5) and in the case of poor logistics (truck transport < 7.5 t) it could take ~40 years until the CO2 balance reaches at least net 0. However, under a RCO2 max scenario, considerable CO2 drawdown could be reached even with a relatively large particle size. It can be argued that RCO2max is high and has not been proposed by any of the other studies in context with basalt application. The exact mechanisms of CO2 drawdown during basalt weathering and the corresponding RCO2max are still unclear (see above) but in any case the assumed drawdown is close to the drawdown for e.g. olivine, showing the potential within ultramafic rocks in general. We note that grinding down to < 1 μm is to our knowledge technically not possible at this moment at industrial scale. Notwithstanding the practical implications of the high energy demand (see section 3.3) it goes to show however that the smallest technically possible grain size should be aimed at. The comminution of material to a grainsize in the range of 10 μm could be achieved with lower energy input, than assumed in prior publications, given the application of state of the art milling equipment (de Bakker, 2014;Gao et al., 2002).

Cumulative CO2 removal
Assuming the maximum application on both intensive grassland (0,576 M ha) and agricultural land (1,326 M ha), we can estimate the amount of CO2 drawdown potential for the application of 10 kg rock/m² (100t/ha). The cumulative amount of basalt applied in this case is 190 Mt. Using rail transport and RCO2 low, this amounts to 4.1 and 14.8 Mt CO2 for the < 100 and < 10 μm fraction, respectively. The annual output of greenhouse gases in Austria amount to roughly 80 Mt CO2e (Klimaschutzreport, 2019).This translates into a net removal of 5.2 % and 18.5 % of Austria's annual greenhouse gas emissions (CO2e) within the observed ~8 year cycle (see table S2). It is important to note however, that current transportation practices could easily turn this into a positive emission technology (see figure   4a and table S2). For RCO2max the potential drawdown would be tremendous from 25 % (< 100 μm) up to a maximum of 86.6 % (< 10 μm) of Austria's annual greenhouse gas emissions (CO2e) within a ~8 year cycle. This is in good agreement with the study of (Moosdorf et al., 2014), who used a comparable RCO2 to evaluate the potential of ultramafic rocks and found that on a global scale transport would not influence CO2 drawdown in way which makes the application unfeasible. We note however that including reaction rates, the net CO2 removal after ~8 years in our estimate is 105 (< 100 μm) and 364 (< 10 μm) kg CO2 t -1 . The range of 500 -1000 kg CO2 t -1 obtained for complete dissolution is misleading because it hides the timeframe needed to achieve such drawdown.

Associated energy requirement
The electricity generation in Austria amounted up to 73460 GWh (20901 GWh of that was generated through biogenic or fossil fuels) in the year 2019 (https://www.e-control.at/betriebsstatistik2019) in contrast to 32904 GWh and 3595 GWh, needed for grinding down of 190 Mt of basalt to < 10 μm and < 100 μm, respectively. Approximately 45% or 4.9 % of the annual power generation is required.
Given the relatively long time for complete dissolution of the powder, it can be argued, that application would not be carried out every year and the optimal application rate is yet to be determined. Still, even if the application cycle is 10 years the energy requirement is 4.5% (< 10 μm) 0.5%(< 100 μm) of annual electricity generation. A detailed analysis of the transformation and growth potential of the energy sector in Austria over the next decades is outside the scope of this study. However we note, that back until the year 2001, the amount of electricity import was always greater than export. Reducing energy demand is considered one of the key factors in fighting climate change over the next decades and in this context liberating 0.5% of the annual electricity generation for rock comminution will not be an easy task.

The amount of basalt needed
With an assumed density of 3g/cm³, volumes of volcanic rock might be found beneath the surface (Heritsch, 1982). 100 t ha -1 is high and it has been shown that considerable CO2 drawdown can also be achieved with lower application rates (Dietzen et al., 2018;ten Berge et al., 2012), potentially allowing to reduce the amount of basalt. In any case, the mafic lavas of the central European volcanic province in Germany comprise several thousand km³ (Jung et al., 2005) and the eruptive volume of the Vogelsberg (central Germany) alone is ~600 km³ (Bogaard et al., 2003). (Kereszturi et al., 2011) calculated a volume of volcanic rocks for the Bakony-Balaton Highland Volcanic Field in the Pannonian Carpathian region of ~2.9 km³ and large volumes of volcanic rocks are present in the bohemian massif, for instance in the doupov mountains (123 km³) (Shrbený, 1995). The chaine de puys erupted ~10 km3 of basalt and trachyte (Martel et al., 2013). Conclusively, there is potential for basaltic rock mining in central Europe for several application cycles on agricultural land, not just in Austria. However, mining even if related to sustainable agriculture and climate change mitigation inevitable is in confrontation with other aspect of environmental protection, related to biodiversity, protected areas and economic and societal factors, related to tourism and recreational value. The enormous touristic potential of some of the regions, by means of cultural heritage, the thermal spa tourism often related to Cenozoic volcanism, agricultural use in general and the high quality wines often related to volcanic terrains in particular put limits on mining capacity of the regions. In addition, environmental impact assessment of related mining projects can be enduring. Even if a first assessment shows the enormous volumes of volcanic rocks in Central Europe, not every occurrence will provide the desired chemical composition (see figure 2) and high quality outcrops are already being heavily exploited. Therefore if basalt fertilizer is to play a more prominent role in Austria's (and Europe's) land management practices, strategic planning of the resources use is now warranted on the European level.

Conclusion
This study assessed the potential of EW through application of basalt on agricultural area in Austria.
Two alternative scenarios for basalt with high (RCO2_max = 880 kg CO2 t -1 ) and low (RCO2_low = 282 kg CO2 t -1 ) CO2 drawdown potential were evaluated for a single application and a reaction time of ~8 years and three different grain sizes (< 100 μm, < 10 μm and < 1 μm). For a transport distance of 300 km the net drawdown of CO2 increases with decreasing grain size from 22 kg CO2 t -1 to 125 kg CO2 t -1 in the RCO2_low scenario, if transport is carried out by railway. Our results also indicate, that road transport would turn the application of a < 100 μm grain size ineffective or worse, even contribute CO2 emission. However, transport becomes less important with decreasing grain size. Application of of 100 t ha -1 with a < 10 μm grain size could draw down approximately 2% of Austria's annual greenhouse gas emissions. The huge energy demand related to the grinding of this amount of rock, based on a application cycle of 10 years, would require up to ~5 % of Austria's total annual power generation. Larger grain sizes have been proposed to circumvent this problem, but our results suggest this energy will have to be available, if enhanced basalt weathering is to make an impact on the CO2 budget.
However, at the moment, uncertainties related to effective mechanisms of CO2 consumption during weathering hinder precise quantification. Stoichiometric dissolution of basalt including Al 3+ , could withdraw higher amounts of CO2 (RCO2_max = 880 kg CO2 t -1 ), in which case the application of larger grain sizes (< 100 μm) could be sufficient. Uncertainties regarding the actual field weathering rates of basalt powder in soil further complicate our ability to quantify the CO2 drawdown related to EW. The precise determination of these rates remains a crucial requirement towards the large scale application of EW. Further studies from laboratory to field scale under different conditions will be necessary to overcome these deficiencies to put the available resources to optimal use within a European strategy. In addition the influence of agricultural practice on greenhouse gas emissions is increasingly put into focus in the context of climate policy, e.g. the EU LULUCF regulation (Romppanen, 2020), and soil amelioration through the application of basalt could become an integral part of soil carbon management.
At last, the energy demand related to EW produces an obvious conundrum. EW is in conflict with the need to reduce energy consumption and the necessary phasing out of fossil fuel based energy. As such, EW relies on our transformation into a sustainable society with a low carbon economy and abundant low carbon power supply to contribute to climate change mitigation. It does vice versa not contribute to this transformation.

Acknowledgements
We thank Sylke Hilberg for insightful discussions during the preparation of this article. Diego Bedoya-González and Timo Kessler are thanked for their help with creation of the map. Kayla Iacovino is thanked for providing an excel sheet for plotting TAS diagrams on her website (www.kaylaiacovino.com). This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Bibliography
Adrees  100 μm grain < 100 μm powder Figure S1: Time for dissolution of a 100 μm basalt grain, calculated using the shrinking core model. Time for 97.5 % dissolution of the grain is 1205 years. According to the chosen grain size distribution for the < 100 μm basalt powder, 99,9 % of the powder will dissolve within this timeframe.