Data set of dissolved major and trace elements from the lacustrine systems of Clearwater Mesa, Antarctica

This article presents analytical observations on physicochemical parameters and major and trace element concentrations of water, ice, and sediment samples from the lake systems of Clearwater Mesa (CWM), northeast Antarctic Peninsula. Geochemical analyses include inductively coupled plasma mass spectrometry (ICP-MS) for cations and trace elements and ion chromatography for anions. Some figures are included (i.e. Piper and Gibbs diagrams) which indicate water classification type and rock-water interactions in CWM, respectively. It also contains PHREEQC software output, listing the chemical speciation for dissolved elements, Saturation Indexes (SI), and modelling outputs. Each lake SI are also illustrated in a figure. Finally, total organic and inorganic carbon (TOC and TIC, respectively) were determined for bottom lake sediments and marginal salt samples. This information will be useful for future research assessing the impacts of anthropogenic pollution and the effects of climate change, providing insights into naturally occurring geochemical processes in a pristine environment, and evaluating geochemical behaviour of dissolved elements in high-latitude hydrological systems. These data correspond to the research article “Dissolved major and trace geochemical dynamics in Antarctic Lacustrine Systems” [1].

and marginal salt samples. This information will be useful for future research assessing the impacts of anthropogenic pollution and the effects of climate change, providing insights into naturally occurring geochemical processes in a pristine environment, and evaluating geochemical behaviour of dissolved elements in high-latitude hydrological systems. These data correspond to the research article "Dissolved major and trace geochemical dynamics in Antarctic Lacustrine Systems" [1].
© 2020 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY license.

Specifications table Subject Earth and Planetary Sciences Specific subject area
Earth-Surface Geochemical Processes Type of data Table  Figure Other (Software Output) How data were acquired Major cations and trace elements were determined by inductively coupled plasma mass spectrometry (PerkinElmer Sciex ELAN 90 0 0 ICP/MS, PerkinElmer Nexion, Thermo icapQ or Agilent 7700, ActLabs, Canada). Chloride and sulphate were determined by chemically suppressed ion chromatography with conductivity detection (Thermo, model Constametric 3500, with Dionex suppressor and IonPac AS22 Dionex column -4 × 250 mm-for anions, CICTERRA-UNC, Argentina). In situ measurements include: alkalinity (volumetric methods), pH and Eh (Hach digital detector), electrical conductivity, temperature, and total dissolved solids (TDS, Hach conductivimeter). TIC and TOC values were determined in sediment samples following Loss on ignition method (LOI, Heiri et al. 2001). Chemical information was processed with PHREEQC Interactive software (U.S. Geological Survey) [2] . Sediment texture was determined with a particle analyser (Horiba LA-950, LabGEO, CICTERRA-UNC, Argentina). Data format Raw analyzed Other (Software Output) Parameters for data collection Samples were collected following standard protocols [3] . Parameters for data collection are: 1-Temperature, pH, redox potential, alkalinity, electrical conductivity, and TDS measured in situ .

Value of the data
• Given the logistical constraints associated with collecting information from this unique and scarcely known high latitude environment, these observations present an opportunity for hydrological and geochemical research without the need for resampling. • This information can be helpful for multidisciplinary researchers, such as environmentalists, biologists, physics, and geologists, and can be readily incorporated into meta-analyses. • This information can be used to identify and analyse processes controlling high latitude lake chemistry through geochemical diagrams and models, and compare them with other environments. • The present information contributes to a growing database that can serve as a baseline in environmental change and pollution studies (physicochemical and trace element) for Antarctic systems.

Data description
Water, ice, and sediment samples from 35 different lakes and ponds were collected during the austral summer (15-29 January 2015) on Clearwater Mesa (CWM), located on James Ross Island, northeast Antarctic Peninsula [1] . Fig. 1 a shows the different hydrogeochemical environ- ments defined according to the relationship between pH vs. Eh (adapted from [4] ). Water ionic classification is shown in Fig. 1 b with a Piper diagram [5] .
Physico-chemical parameters and major ion concentrations were determined for 37 lake and ice samples, and are presented in Table 1 . This table also indicates lake area; Mg 2 + /Ca 2 + and Na + /(Na + + Ca 2 + ) ratios; and sediment TOC and TIC. Fig. 2 shows TDS values vs. Na + /(Na + + Ca 2 + ) ratios in a Gibbs diagram [6] . Table 2 displays the hydrochemical statistical values for each hydrogeological system.  Table 3 presents dissolved trace element concentrations. They were normalized to the Upper Continental Crust [7] and a spidergram is shown in Fig. 3 , where World average [8] and the mean James Ross Island Volcanic Group geochemical composition have been indicated [9] . Table  4 shows statistical values for differentiated hydrogeological system's trace elements.
PHREEQC models are presented in the Software Output, which is included in the Supplementary Material . It comprises ion speciation, Saturation Indexes, and geochemical models (i.e. mixing modelling between Blancmange Glacier and Cecilia samples and inverse modelling between Blancmange Glacier and Florencia samples). Finally, Fig. 4 illustrates those minerals that are prone to precipitate due to their positive Saturation Index (SI).

Sampling and analyses
Water and sediment samples were collected from 35 different lakes and ponds on Clearwater Mesa during the 2015 Antarctic summer field campaign following standard protocols [3] . Two  Table 2 Statistical values for each hydrogeological system's physical and chemical characteristics.   ice samples (i.e., Blancmange Glacier and Lake Natasha-ice), were also collected, along with a precipitated salt sample from the margin of Lake Andrea. Water temperature, pH, redox potential, electrical conductivity, TDS, and alkalinity were measured at each waterbody in situ . Redox potential and pH were measured with a Hach digital detector, while temperature, TDS, and electrical conductivity were measured using a digital Hach conductivimeter. Alkalinity was measured as CaCO 3 by end point titration in the field, and using a 0.16 N H 2 SO 4 solution until pH = 4.5. For the determination of anions, major cations, and trace elements, samples were vacuum-filtered in the field with 0.22 μm pore size cellulose filters (HA-type, Millipore Corp.). An aliquot was stored in polyethylene bottles at 4 °C for the determination of chloride and sulphate by chemically suppressed ion chromatography with conductivity detection. The other aliquot was acidified (pH < 2) with concentrated, redistilled and ultrapure HNO 3 (Sigma-Aldrich) for the analytical determination of major and trace elements by inductively coupled plasma-mass spectrometry (ICP-MS, Activation Laboratories Ltd., Ancaster, Ontario, Canada). Major and trace elements concentrations were validated using NIST (National Institute of Standards and Technology) 1640 and Riverine Water Reference Materials for Trace Metals certified by the National Research Council of Canada (SRLS-4), and detection limits are reported in the corresponding tables. Statistical parameters (i.e. mean and median) were calculated for each hydrogeological system's trace elements and physical and chemical characteristics.

Sediment analysis
Bottom lake sediment samples were collected from 32 shallow lakes using a standard scoop sampler. Lake Katerina was sampled 3 times in order to characterize spatial variability in this   large lake. Particular attention was paid to collect the sediment/water interface. Samples were skimmed from the benthic surface (top 2-3 cm), stored in plastic bags, and refrigerated until analysis. TOC and TIC was determined in sediment samples using the loss on ignition method (LOI; [10] ). Prior to their analysis by LOI, they were partitioned and lyophilized in order to remove water by sublimation. To prepare the samples, sediment was first ground to obtain a homogeneous sample. A quantity of 2 g was then weighed on a high-precision analytical balance (Precisa XR 2058-DR) and placed in numbered crucibles. Each sample was weighed to obtain the initial weight. The organic matter content was determined by calculating the difference in mass between sediment samples dried to a constant weight at 105 °C, and after furnacing at 550 °C for 4 h. Carbonate content was calculated as the mass loss after burning the LOI residue at 950 °C for 2 h [10] .

Geochemical modelling
Chemical information was processed with PHREEQC [2] constructed using the AQUACHEM PHREEQC interface. These programs were used to calculate the SI for different mineral phases, element speciation in the given conditions (i.e. the element distribution in all the possible dissolved chemical species that can be found in the samples), and geochemical weathering and mixing models. Software outputs are presented in the Supplementary Material . The SI is the ratio between the ion activity product for the given material and the reaction constant at a given temperature. When SI > 0, it means that the solution is supersaturated with respect to the mineral phase and may therefore precipitate, whereas when SI < 0 the solution is below saturation of the specified mineral. Geochemical modelling uses the interaction of cations and anions as a function of temperature, redox potential, pH, and ionic strength. Inverse modelling was performed to quantify weathering processes occurring, first between connected lakes Esther and Tatana, and second between Blancmange Glacier and Lake Florencia. Inverse modelling begins with known initial and final solutions, and available mineral phases which attempts to quantify the processes that lead to the final solution chemistry (e.g., [11] ). In the process, all possible combinations of dissolution and/or precipitation reactions explaining chemical changes observed between the two solutions and the mineral phases are reconstructed [12] . Models were performed under equilibrium with O 2 (g) given that these are surficial hydrological systems. Finally, three mixing models with different proportions of Blancmange Glacier and Lake Cecilia solutions (i.e., one model where both the glacier and lake contribute 50% of the water, one where 20% of the water is from the glacier and 80% is from Lake Cecilia, and one where 80% of the water is from the glacier and 20% is from Lake Cecilia) were made to verify eventual mixing processes.