The migration of total dissolved solids during natural freezing process in Ulansuhai Lake

Journal of Arid Land2012, 4(1): 85 -94doi: 10.3724/SP.J.1227.2012.00085Science Pressjal.xjegi.com; www.chinasciencejournal.comThe migration of total dissolved solids during naturalfreezing process in Ulansuhai LakeYan ZHANG', ChangYou L1", XiaoHong SHI', Chao L1,2'Water Conservancy and Civil Engineering College, Inner Mongolia Agricultural Universit, Huhhot 010018, China;2 Department of Civil and Environmental Engineering, Clarkson University, Potsdam, NY 13699-5710, USAAbstract: High total dissolved solids (TDS) content is one of the most important pollution contributors in lakes inarid and semiarid areas. Ulansuhai Lake, located in Urad Qianqi, Inner Mongolia, China, was selected as the objectof study. Temperatures and TDS contents of both ice and under-ice water were collected together withcorresponding ice thickness. TDS profiles were drawn to show the distribution of TDS and to describe TDS migra-tion. The results showed that about 80% (that is 3.602*108 kg) of TDS migrated from ice to water during the wholegrowth period of ice. Within ice layer, TDS migration only occurred during initial ice-on period, and then perished.The TDS in ice decreased with increasing ice thickness, fllowing a negative exponential-like trend. Within un-der-ice water, the TDS migrated from ice-water interface to the entire water column under the effect of concentra-tion gradient until the water TDS content was uniform. In winter, 6.044x10' kg (16.78% of total TDS) TDS migratedfrom water to sediment, which indicated that winter is the best time for dredging sediment. The migration effectgives rise to TDS concentration in underice water and sediment that is likely to affect ecosystem and water qualityof the Yellow River. The trend of transfer flux of ice water and water- sediment interfaces is similar to that of icegrowth rate, which reveals that ice growth rate is one of the determinants of TDS migration. The process andmechanism of TDS migration can be referenced by research on other lakes with similar TDS content in cold andarid areas.Keywords: arid and semiarid areas; Ulansuhai Lake; total dissolved solids (TDS) migration; natural freezing process; transferflux; ice growth rateLakes in arid and semiarid areas are not only anal, 201 1) than to pollution characteristics of ice-cov-indicator of climate change, but also an important linkered lakes. Mirono and Terzhevik (2000) concludedof water cycle. They have a great effect on ecologicalthat the conductivity of the lake ice was 10%- -20% ofenvironment change (Li et al, 2007). High total dis-that of the lake water in Lake Paajarvi. Huang et al.solved solids (TDS) content is one of the most impor-(2009) described the distribution patterm of nutrientstant pollution contributors because of lttle precipita-and phytoplankton during the icebound season intion, insufficient supply of surface runoff and com-Changchun, China. Roger and Gregory (2009) showedparatively great evaporation intensity in these areasthe effect of pollutant exclusion from lake ice on sea-(Tao, 2001; Wang and Sun, 2007). TDS characteris-sonal circulation. However, TDS migration and col-tics during ice-off period have long been a focus ofmn distribution patterm during ice-on period are notresearch, which mostly concentrated on distribution,well known, and have received lttle attention frommigration and environmental effects (Gupta andresearchers, leading to lacked studies on pollutionDeshpande, 2003; Zhou et al, 2007; Han et al, 2009).mechanism and TDS process in winter.In general, the studies on lake ice paid more attentionDue to man中国煤化工intensity andto the process of ice growth process (Fang et al, 1996;:MYHCNMH GMirono et al, 2002; liescu and Baker, 2007; Jiang et*Corresponding author: ChangYou LI (E mail: ndlichangyou@163.com)36JOURNAL OF ARID LANDVol. 4.heat exchange, the characteristics of TDS pollutionhe largest lake at the same latitude on the earth. Induring the ice-on period may be different from thoserecent years, the water supply to Ulansuhai Lake hasin the ice-free period (Belzile et al, 2002; Lu et al,been decreasing due to climate change and the irriga-2010). Ulansuhai Lake, which is a typical representa-tion quota adopted in Hetao irrigation area, and thetive of lakes in arid and semiarid areas, was selectedwater level has been continuously falling. Meanwhile,as the subject of study to measure TDS content ofTDS pollution has been worsening and industrialboth ice and under-ice water as ice thickened. Thewastewater of high TDS content was drained into thestudy focused on the following two questions: (1)lake (Ren et al, 2008).How and in what volume did TDS migrate amongRegional meteorologic statistics in Urad Qianqiice-water-sediment system during the whole freeze-upshow that in the region of Ulansuhai Lake the meanperiod? (2) Why did TDS migrate and what did theannual air temperature is 6.6°C. The warmest month ismigration imply? The results of this work can be ref-July (24.6°C) and the coldest is January (-10.2°C).erenced by research on other lakes with similar TDSThe average air temperature from November to Feb-content in cold and arid areas.ruary is below 0°C. Ice season lasts for 5- 6 months inUlansuhai Lake, from November/December to April,1 Materials and methodsand the maximum annual ice thickness ranges from 401.1 Study areato 70 cm. In winter, after ice formation, the lake turnsinto a calm water body in which water turbulence andUlansuhai Lake (40936'-41903N, 10843-108*57E)circulation activities are weak. The annual precipita-is located in Urad Qianqi, Inner Mongolia, China andtion is 219.5 mm and 67.9% of it falls between Julyit is part of Hetao irigation system which takes aboutand September. The anual evaporation is 1,396.3 mm5x108 m3 of water from the Yellow River. A complexby E601 evaporator and 47.4% of it occurs betweensource of water, including domestic and industrialMay and July.wastewater from neighboring areas, is drained into thelake Fig. 1). The total area of Ulansuhai Lake is about1.2 Sampling333.48 km2, of which 161.631 km2 is covered withThe station S (Fig. 1) was selected as the samplingreed, and the area of open water is 104.824 km2. It ispoint for its special geographic location. It bydrologi-Ulansuhai lakeWuryuan41°NLinhcd Yellow RiverDengkou米050km108°E中国煤化工Flg.1 Geographical location of Ulansuhai Lake and samplingJYHCNMHG.No.1Yan ZHANG et al: The migration of total dssolved solids during natural freczing process in Ulansuhai Lake8(bcally connects with the rest of the lake in wet seasons3 Ruler .between May and October, i.e. the lake water ex-(achanges between station S and major parts of the lake,while exchanges ceased in dry season during the study,L. Batterywhich impeded TDS exchange between them. Thebottom of station S was flat, and the depth of waterILwas 1.5 m on average.Field observation was arranged during ice-on pe-WateLriod at station S. Sampling was made every two daysfrom Nov.2010 to Apr. 2011 using an ice drill and anThermal lincice saw. After extraction from the ice sheet, the iceFig. 2 lce thickness measurement by thermal line. (a) Status ofcore was quickly removed from the barrel and placedmeasuring; (b) Status of readingon a cutting board. The core was then cut every 10 cmfrom the top using a cut-off saw, which is mounted atbased on an assumption that the lake is a complete andthe end of the cutting board. The ice sample was nextclosed system consisting of atmosphere, ice, water andplaced in a sealed plastic bag or container for laterlake bed; another is to derive a semi -empirical formulamelting and TDS measurement. After logging the icecalled Zubov model according to meteorological data,samples, water samples were taken through smallwhich is simple and accurate for application in smallholes using a 1,000 mL self-made“syringe”sampler.areas (Li and Riska, 2002).To avoid the disturbance of thermal exchange andIn this study, we built a Zubov model based on Cu-solar radiation, the snow on the ice surface wasmulative freezing degree-day and corresponding icecleaned immediately after snowing.thickness (Fig. 3). Through fitting the curve, the em-13 Data colectionpirical formula of Ulansuhai between ice thicknessTemperature data were collected by a Onset Hoboand cumulative freezing degree- day was established:Temp Pros with an accuracy of 0.259C. TemperatureFDD=0.002r2 -0.2251T.(1)was recorded every 5 min and uploaded each month.The TDS content data were obtained with an EC60Where T is ice thickness (cm); FDD is cumulativeTDS meter at 25°C every two days in freezing process. freezing degree-day (°C). .The density of ice was measured using quality-volumeCumnultive freing degree-day (C)method. The ice thickness was measured by a_20000_60self- made thermal line, as depicted in Fig. 2 (Li et al,-1R =0.98872005). The full length L1 of the thermal line wasmeasured before observation, and then the thermal line县-20was heated by battery to melt the ice surrounding it暑-30,until it could move freely in vertical direction. Thethermal line was next pulled until the baffle plate wasblocked off by the down surface of ice. The partiallength Lr of the thermal line above the upper surface-6was next measured. The differential value ofLI and L2Fig. 3 Relationship between ice thickness and cumulativewas the ice thickness.freezing degree-day2 Results and discussion2.2 TDS migration2.1 Ice growth2.2.1 TDS σ中国煤化IvaterThere are two methods to calculate ice thickness after The TDS cont:YHC N MH Ging ice-con pe.continuous ice sheet forms: one is to build a growthriod was much higher than tnat Derore ice-on period,model through analyzing the heat exchange process while TDS content of ice was much lower than that of.88JOURNAL OF ARID LANDVol. 4water in ice-on period (Fig. 4). The phenomenon indi-the above theory of crystallography is not consistentcates that TDS migrated from ice to under-ice water inwith the result that the ice also contained TDS thoughfreeze-up process, which resulted in TDS exclusion.its content was lower than that of water, as describedBased on calculation, about 80% of the TDS was re-in Fig. 4. The reason lies in that much TDS betweenjected from lake ice at an ice thickness of 59 cm. Thefresh ice platelets failed to migrate to water and waseffect of TDS exclusion can be explained by two theo-entrapped during the ice growth process (Fig. 6). Theries known as crystallography theory and liquid-solidentrapped procedure of TDS is similar to brine expul-phase equilibrium theory.sion from sea water (Weeks and Ackley, 1989). ManyWhen lake water freezed, small, flat platelets offactors affect the volume of TDS entrapped in freshpure ice formed firstly. As ice mass grew, the plateletsice platelets: ice growth rate, TDS concentration incommingled and bonded to form a compliant slush.water, ice structure and under ice turbulence. WhenUnder sustained freezing temperatures, the platelettemperature is extremely low, growth rate acceleratesmass thickened and grew together, forming ansharply, and more TDS is incorporated in ice at theever-stiffening ice cover. Near the ice-water inter-bottom of ice sheet (Weeks and Lee, 1958; Nakawoface, water molecules were precipitated with the in-and Sinha, 1981). With ice thickness increases, growthteraction of hydrogen bonding and adhered to therate slows because of the decreasing rate at which heatdown surface of ice, and then formed ice. At the sameis transferred from the bottom of ice to atmospheretime, TDS would be removed from ice to water be-(Maykut, 1986). Thus less TDS was entrapped in icecause of the effect of TDS exclusion from lake ice,as ice grew, which is consistent with what Fig. 4 (a)H(f)which explains the fact that the concentration of waterdescribe that TDS concentration in ice decreases as icemolecules is much lower than that of the whole liquidsheet thickens.and the concentration of TDS in water is much higherThe effect of TDS exclusion can also be explainedthan that nearby the ice-water interface. With theby liquid-solid phase equilibrium theory. DABE is adriving force of concentration gradient, the waterfreezing curve of lake water in Fig. 7. D is the freezingmolecules in liquid were diffused to ice-water inter-point for pure water. If lake water is cooled from theface, while TDS nearby the liquid solid interface wereinitial temperature To to TA, water begins to freeze (Ifdifused to liquid, just as depicted in Fig. 5. However,the TDS content of lake water increases, the freezingTDS (g/L)TDS (gL)TDS (g[L)TDS (