Geo log i cal Quar terly, 2017, 61 (2): 480–490 DOI: http://dx.doi.org/10.7306/gq.1349
New lu mi nes cence ages re veal Early to Mid dle Weichselian de pos its in cen tral Lat via
Kristaps LAMSTERS1, *, Edyta KALIÑSKA-NARTIŠA2, Vitªlijs ZELÈS1 and Hel ena ALEXANDERSON2
1 Uni ver sity of Lat via, Fac ulty of Ge og ra phy and Earth Sci ences, Jelgavas Street 1, LV-1004, Riga, Lat via
2 Lund Uni ver sity, De part ment of Ge ol ogy, Sölvegatan 12, 223 62 Lund, Swe den
Lamsters, K., Kaliñska-Nartiša, E., Zelès, V., Alexanderson, H., 2017. New lu mi nes cence ages re veal Early to Mid dle Weichselian de pos its in cen tral Lat via. Geo log i cal Quar terly, 61 (2): 480–490, doi: 10.7306/gq.1349
New op ti cally stim u lated lu mi nes cence (OSL) ages show that sandy de pos its over lain by Late Weichselian subglacial till in cen tral Lat via are of Early to Mid dle Weichselian age. The finer chro no log i cal res o lu tion of un con sol i dated sed i ment de po si - tion in the Cen tral Lat vian Low land (CLL) re mains rel a tively un stud ied, and here we pro vide a first char ac teri sa tion of the de - pos its with re spect to their age. Three OSL ages rang ing be tween 84 ± 9 ka and 112 ± 11 ka sug gest that the de pos its stud ied in the CLL are of Early Weichselian age (MIS 5). We found no Mid dle Weichselian de pos its in the CLL, and as sume that any such youn ger sed i ments might have been eroded dur ing the ad vance of the Zemgale Lobe in the Late Weichselian. One site, in the ice-mar ginal zone ad ja cent to the interlobate area, has nev er the less de pos its dated to 44 ± 10 ka cor re spond ing to the Mid dle Weichselian (MIS 3). Our re sults are com pat i ble with ex ist ing ESR ages on three sets of Portlandia arctica shells from the cen tral part of the low land; the shells had been in cor po rated into gla cial de pos its dur ing later gla cial ad vances. Fi nally, our find ings largely sup port ice-free con di tions dur ing the Early and Mid dle Weichselian in the mid dle and south ern part of cen tral Lat via.
Key words: op ti cally stim u lated lu mi nes cence (OSL) dat ing, subglacial bedforms, Early Weichselian, Mid dle Weichselian, cen tral Lat via.
INTRODUCTION
The chro nol ogy of the Weichselian Gla ci ation in cen tral Lat - via is not well es tab lished due to the lack of re li able ab so lute dat ing. No op ti cally stim u lated lu mi nes cence (OSL) dates of Early and Mid dle Weichselian age have un til now been avail - able for cen tral Lat via (cf. Zelès et al., 2011), and only a few OSL ages re flect ing Mid dle Weichselian time have been re - ported from other lo cal i ties in Lat via (Zelès et al., 2011; Saks et al., 2012). Some ra dio car bon dates are known but they are ques tion able be cause of pos si ble con tam i na tion (Dreimanis and Zelès, 1995). Elec tron spin res o nance (ESR) ages of Portlandia arctica shells re de pos ited in gla cial sed i ments of the Cen tral Lat vian Low land (CLL) dur ing gla cial ad vances range from 86.0 ± 6.8 ka (till) to 105.0 ± 9.2 ka (gravel) (Molodkov et al., 1998). The deglaciation his tory has also been in ves ti gated by cosmogenic 10Be age de ter mi na tion (Rinterknecht et al., 2006, 2008) but this method suf fers from in suf fi cient res o lu tion that ren ders dif fi cult ac cu rate cor re la tion and sep a ra tion of ice-mar ginal for ma tions.
OSL dat ing pres ents an op por tu nity to date the dom i nant con stit u ents of the de pos its them selves (Thrasher et al., 2009),
par tic u larly of those that are sandy. These de pos its, there fore, con sti tute suit able ma te rial for chro no log i cal stud ies. The OSL method has been fre quently used in the last de cade to ob tain the age of glacioaquatic, flu vial, lac us trine and ae olian de pos its in the Bal tic re gion (Kalm, 2006; Molodkov et al., 2010; Rattas et al., 2010; Raukas et al., 2010; Zelès et al., 2011; Saks et al., 2012; Satkñnas and GrigienÅ, 2012; Baltrñnas et al., 2013;
Kaliñska-Nartiëa et al., 2015a, b, 2016). De spite this, ques tion - able OSL ages have been fre quently ob tained from glaciolacustrine and glaciofluvial de pos its (e.g., Raukas et al., 2010). A pos si ble cause for this is in com plete bleach ing of sand grains due to rapid de po si tion, tur bu lent streams, sed i men ta - tion in dark ness or sig nif i cant wa ter tur bid ity (cf. Weckwerth et al., 2013). The re sults of pre vi ous in ves ti ga tions of Pleis to cene chro nol ogy and par tic u larly of the deglaciation of the last gla ci - ation (GuobytÅ, 2004; Zelès and Markots, 2004; Kalm, 2006, 2012; Rinterknecht et al., 2006, 2008; Raukas et al., 2010;
GuobytÅ and Satkñnas, 2011; Zelès et al., 2011; Lasberg and Kalm, 2013) in di cate un re solved strati graphic is sues, such as the lack of the suf fi ciently ac cu rate ages of Late Pleis to cene de - pos its, par tic u larly of Early and Mid dle Weichselian age.
To im prove the chro nol ogy of the Weichselian Gla ci ation in cen tral Lat via, we dated the sands over lain by Late Weichselian till with OSL. Mostly, these sands are found within the cores of subglacial bedforms (de po si tion of sandy sed i ments pre ceded the for ma tion of subglacial bedforms) in the area of the Zemgale Ice Lobe (Lamsters and Zelès, 2015), and one sam ple co mes from an interlobate area; these de pos its have not been pre vi ously in ves ti gated by lu mi nes cence meth ods. The is sues
* Corresponding author, e-mail: kristaps.lamsters@gmail.com Received: July 9, 2016; accepted January 19, 2017; first published online: March 9, 2017
con cern ing OSL dat ing of ques tion ably well-bleached sed i - ments as well as the chro no log i cal con se quences of the ages are dis cussed.
STUDY AREA AND STRATIGRAPHY
GEOLOGICAL SETTING
Our study area is lo cated in the cen tral part of Lat via (Fig. 1) com pris ing the CLL, and it lies within the in ner pe riph eral zone of the last Fennoscandian Ice Sheet (Straume, 1979; boltiÖë, 1989; Zelès and Markots, 2004). The CLL is a glaciodepression of di ver gent type, which is linked with the cen tral Lat vian bed - rock de pres sion (Zelès, 1993). The CLL ex pe ri enced con sid er - able gla cial ero sion through the Pleis to cene glaciations that re - sulted in a rel a tively thin Qua ter nary suc ces sion – on av er age of 10–20 m and up to 20–40 m only be low the high est gla cial land - forms (Juškeviès, 2000, 2001; Meirons, 2002). The de pos its are thin ner in the cen tral part of the CLL, <10 m and, in some places, only a few metres thick (Juškeviès, 2000, 2001;
Meirons, 2002; Zelès and Markots, 2004).
The gla cial landsystems in the CLL were gen er ally formed dur ing the North (Linkuva in Lat via) and Mid dle Lith u a nian (Gulbene in Lat via) gla cial phases of the Late Weichselian Gla - ci ation, when the Zemgale Ice Lobe of the Riga Ice Stream (boltiÖë et al., 1977) was re-ac ti vated dur ing the deglaciation.
The di ver gent flow char ac ter of the Zemgale Ice Lobe is dem -
on strated by the ori en ta tion of stream lined bedforms (Lamsters and Zelès, 2015). The pres ent-day to pog ra phy is dom i nated by ra dial and trans verse subglacial bedforms – mostly drum lins and ribbed mo raines (Zelès, 1993; Lamsters, 2012; Lamsters and Zelès, 2015).
The bed rock in the CLL is com posed of Up per De vo nian Frasnian and Famennian sed i men tary rocks: sand stone, siltstone, clay, do lo mite, marl and gyp sum, as well as of Mid dle De vo nian terrigenous rocks in its NE part (Brangulis et al., 1998). The youn gest Perm ian, Tri as sic and Ju ras sic terrigenous and car bon ate rocks are dis trib uted in the SW cor - ner of the CLL (Brangulis et al., 1998). The bed rock sur face, in gen eral, dips to wards the N in the di rec tion of the Gulf of Riga.
The al ti tude of bed rock sur face var ies from 110 m a.s.l. to 60 m b.s.l. at the south ern end of the Gulf of Riga.
QUATERNARY DEPOSITS
The Qua ter nary de pos its mainly con sist of het er o ge neous Late Weichselian till interlayered with and un der lain by sandy de pos its (Savvaitovs and Straume, 1963; boltiÖë, 1963;
Ginters, 1978; Juëkeviès, 2000, 2001; Meirons, 2002). The ex - is tence of an older till of Saalian age has been con firmed in: (1) lo cal bed rock de pres sions and palaeoincisions (boltiÖë, 1963;
Juškeviès, 2000, 2001; Meirons, 2002), (2) glaciotectonic land - forms be low the youn gest till beds (Dreimanis, 1935; Dreimanis and Zelès, 1998, 2004), or (3) as intraclasts in the youn ger tills (Dreimanis, 1935). The petrographic and tex tural com po si tion
Fig. 1. The location of the study area (Central Latvian Lowland) and OSL sampling sites
of till beds tends to be very sim i lar (Mironovs et al., 1962;
boltiÖë, 1963; Ginters, 1978), thus their strati graphic di vi sion is com pli cated and some times dif fi cult to de ter mine. The strati - graphic po si tion of the tills is also poorly un der stood, be cause of the lack of qual i ta tive dat ing re sults. For ex am ple, the up per dark-grey till bed was thought to have been de pos ited dur ing the Saalian (Danilªns, 1973) in West ern Lat via. A re-in ves ti ga - tion in the west ern part of Lat via, how ever, re vealed its Late Weichselian age (Saks et al., 2012). Thin and de formed grey till beds of un known strati graphic po si tion have also been iden ti - fied at sev eral quar ries in the CLL (Lamsters and Zelès, 2015).
A sim i lar grey till has been found at the banks of the Daugava River near Daugmales TomÂni (see Fig. 1), lo cated in a ribbed mo raine, and it was in ter preted as a Saalian till, be cause no Portlandia arctica shells were found (Dreimanis and Zelès, 1995, 1998; Molodkov et al., 1998).
Com monly, the Pleis to cene se quence in the CLL con tains two or more till beds. For ex am ple, two Weichselian till beds interbedded with sorted de pos its have been rec og nized in the area up-gla cier of the North Lith u a nian ice-mar ginal suc ces sion (boltiÖë, 1963; Savvaitovs and Straume, 1963; Dreimanis and Zelès, 1995; Zelès and Markots, 2004). Three till beds have been found in lo cal bed rock de pres sions (Ginters, 1978;
Juëkeviès, 2001), and up to four till beds have been re ported from North Lith u a nia (Gaigalas and Marcinkevièius, 1982;
Baltrñnas et al., 2005); the lower till bed has been at trib uted to the Grñda stadial (Gaigalas, 1995; GuobytÅ, 2004), which co in - cides with the max i mum ex tent of the last Fennoscandian Ice Sheet. The up per till bed cor re lates with the Bal tic stadial of the Late Weichselian Gla ci ation (Gaigalas, 1995; GuobytÅ, 2004).
The high est num ber of lithologically dif fer ent till beds (up to 5) has been found in a bore hole at the west ern side of the Vidzeme Up land (Mironovs et al., 1962). Sim i lar lithological dif - fer ences are linked to glaciotectonically de formed till beds, for ex am ple at up-gla cier slopes of ribbed mo raines. Up to five stacked Late Weichselian till beds un der lain by one or two pos - si ble Saalian till beds are fre quently re ported (boltiÖë, 1989;
Dreimanis and Zelès, 1998, 2004; Lamsters and Zelès, 2015). It is fea si ble that the youn ger till beds, which formed due to subglacial thrust ing, are of the same age. It is well-es tab lished that glaciotectonic de for ma tion played a sig nif i cant role both in drum lin and ribbed mo raine for ma tion and in chang ing the orig i - nal Pleis to cene sed i ment bed ding and thick ness (Levkov, 1980; boltiÖë, 1989; Zelès, 1993). This is why interlayers of sorted de pos its, rafts of pre-Late Pleis to cene de pos its and De - vo nian bed rock, and up wardly in jected clastic dykes are com - mon in the CLL (Dreimanis, 1935, 1992; boltiÖë, 1963; Zelès, 1993), par tic u larly in glaciotectonic land forms, for ex am ple, in ribbed mo raines (Dreimanis and Zelès, 1998, 2004) and on the prox i mal slopes and ad ja cent ar eas of the North Lith u a nian ice-mar ginal suc ces sions (boltiÖë, 1963).
The cover on top of the up per Late Weichselian till bed con - sists pre dom i nantly of 5–6 m thick glaciolacustrine de pos its of the Bal tic Ice Lake and youn ger coastal de pos its in the prox im - ity of the Gulf of Riga and a 2–3 m thick se quence of glaciolacustrine de pos its of the Zemgale ice-dammed lake in the north ern and cen tral parts of the plains of the CLL. Ad di tion - ally, 3–8 m thick glaciolacustrine de pos its of the Daudzeva proglacial lake are dis trib uted fur ther to the south-east of the study area. Sandy ae olian sed i ments form ing in land dunes oc - cur in places (Juškeviès, 2000, 2001; Meirons, 2002).
GEOLOGICAL DESCRIPTION OF OSL SAMPLING SITES
Three lo cal i ties with sandy de pos its over lain by subglacial till of the Late Weichselian Gla ci ation were cho sen to con strain
their de po si tion in time and these are Lªèi, Kalna Bñbli and Nereta (Fig. 1). All these sites are lo cated in subglacial bedforms formed dur ing the de cay of the last Fennoscandian Ice Sheet. Lªèi and Kalna Bñbli are sit u ated in the land forms (area cov ered by the Zemgale Ice Lobe), while Nereta is lo - cated in the glaciotectonised ice-mar ginal suc ces sion of the Mid dle Lith u a nian gla cial phase, which stretches along the west ern edge of the interlobate area be tween the Zemgale and Lubªns Ice Lobes. The de po si tion of sandy sed i ments oc curred prior to for ma tion of subglacial bedforms. The de pos its rep re - sent undeformed cores of subglacial bedforms, which were formed due to ero sion of pre-ex ist ing sorted sed i ments. Later, till was de pos ited and de formed (Lamsters and Zelès, 2015).
The depositional en vi ron ment of the sam pled sandy de pos its is likely flu vial or glaciofluvial: see site de scrip tions be low.
In the Lªèi Quarry, up to 3 m thick red dish-brown subglacial till cov ers fine to me dium-grained sand. The hor i zon tal-lam i - nated sand was sub jected to OSL sam pling at 4.7 m and 9.2 m depth (Fig. 2A, B). The Lªèi Quarry is lo cated in a drum lin (Lamsters and Zelès, 2015), the core of which con sists mainly of fine- and me dium-grained sand. The most fre quent lithofacies are pla nar cross-bed ded, trough cross-bed ded, hor i - zon tal-lam i nated and climb ing rip ple cross-lam i nated sand.
Hor i zon tal-lam i nated sand, which was sam pled, is thought to have been de pos ited from rapid streams in the up per flow re - gime. The com po si tion and tex ture of the sorted de pos its at the Lªèi Quarry and their lithofacies sug gest de po si tion from streams in a flu vial or shal low ba sin en vi ron ment, al though we also can not ex clude glaciolacustrine sed i men ta tion in a ba sin with high sed i ment sup ply and flu vial ac tiv ity.
A sim i lar fine sand se quence was sam pled at a depth of 5 m at Kalna Bñbli (Fig. 2D). The Pleis to cene se quence here con sists of hor i zon tally-lam i nated, pla nar and trough cross-bed ded sandy and oc ca sion ally grav elly de pos its in di - cat ing de po si tion by mod er ately rapid streams. The sorted de - pos its are over lain by Weichselian subglacial till. The till is interbedded with de formed sandy de pos its and is lo cally boudinaged, which re sults in a dis con tin u ous till layer, which is thought to have been de formed subglacially dur ing the for ma - tion of ribbed mo raine. The top most part of the sec tion con - sists of 1.5 m of varved clay that di rectly cov ers the till unit.
This clay is thought to have been de pos ited in a lo cal ba sin af - ter the for ma tion of the ribbed mo raine.
In the Nereta Quarry, sam pling was car ried out in fine sand at 5.5 m depth. Here, the Pleis to cene se quence con sists of a suc ces sion of till and sand lay ers un der lain by mas sive and hor - i zon tally-lam i nated sand. The up per part of Pleis to cene suc ces - sion was de pos ited as thrust sheets in ice-mar ginal con di tions be tween the ad vanc ing Lubªns and Zemgale ice lobes, but the age of de po si tion of the sam pled sand is un known.
METHODOLOGY
All lu mi nes cence dat ing sam ples were col lected from four sites by ham mer ing opaque plas tic tubes into freshly cleaned ex po sure sur faces us ing the Eijkelkamp Sam pler Set for Hard Soils sam pling kit. Sam ples were taken in opaque plas tic tubes to avoid ex po sure to day light while sam pling. Along with the OSL sam ples, sed i ment wa ter con tent was as sessed in the field us ing a ThetaProbe Soil Mois ture Sen sor. Four sed i ment sam ples (two from Lªèi, one from Kalna Bñbli and one from Nereta) were treated un der dark room con di tions at the Lund Lu mi nes cence Lab o ra tory (LLL), Lund, Swe den and a de tailed de scrip tion can be found in the fol low ing subchapters (see lo ca - tion of OSL sam pling sites in Fig. 1). Two sam ples from the 482 Kristaps Lamsters, Edyta Kaliñska-Nartiša, Vitªlijs Zelès and Helena Alexanderson
Krimñnas Quarry were ana lysed at the Finn ish Mu seum of Nat - u ral His tory (FMNH), Hel sinki, Fin land, and since we can pro - vide only a lim ited meth od olog i cal back ground and dataset (Oinonen and Eskola, 2009), they are treated as un pub lished data and are only used for com par i son in the In ter pre ta tion and Dis cus sion.
SAMPLE PREPARATION AND MEASUREMENT
The sam ples in opaque plas tic tubes were opened un der dark room con di tions. The sed i ment at each end of the tubes, which might have been ex posed to day light while sam pling, was re tained to de ter mine the dose rate and the wa ter con tent. Only ma te rial from the in ner part of the tube was taken for De mea - sure ments. The 180–250 µm frac tion was ex tracted by wet siev ing. Af ter that, sam ple prep a ra tion in cluded treat ment with 10% HCl for 30 min. and 10% H2O2 for 15 min. to dis solve car - bon ates and or ganic ma te rial, re spec tively. Heavy liq uid (LST Fast Float) with a den sity of 2.62 g/cm3 was used to sep a rate quartz and feld spars. The ex tracted quartz sam ples were treated with 38% hy dro flu oric acid (HF) for 60 min, and fur ther re-treated with 10% HCl for 40 min. to re move pos si ble flu o ride
con tam i na tion. Fi nally, the pu ri fied quartz sam ples were re-sieved with a 180 µm sieve be fore mea sure ment.
At the LLL an au to mated RisÝ TL/OSL reader DA-20 equipped with a cal i brated 90Sr/90Y beta ra di a tion source (dose rate ~0.15 Gy s–1), blue and in fra red (IR) LED (light emit ting di - ode) lamps, and U-340 glass fil ter 7.5 mm thick was used for mea sure ments (BÝtter-Jensen et al., 2000).
WATER CONTENT AND DOSIMETRY
Wa ter con tent was de ter mined by weigh ing a subsample from the OSL-tube, when taken out from the tube (= nat u ral wa - ter con tent) when sat u rated (af ter 24 h cov ered by wa ter) and when dry (af ter 24 h at 105°C) and fur ther cal cu lat ing the nat u - ral and sat u rated wa ter con tent as weight per cent. The dose rate subsamples were dried, ig nited (24 h at 450°C), ho mogen - ised and cast in wax in a fixed ge om e try. Prior to radionuclide con cen tra tion mea sure ments, the casts were stored for at least three weeks (Murray et al., 1987). A high-res o lu tion gamma spec trom e ter at the Nordic Lab o ra tory for Lu mi nes cence Dat - ing, RisÝ, Den mark, was used. The radionuclide con cen tra tions have been con verted into beta and gamma dose rates fol low ing Olley et al. (1996).
Fig. 2. The sedimentary structures observed at the Lªèi (A, B), Nereta (C) and Kalna Bñbli (D) OSL sampling sites
Dmm – diamicton, ma trix sup ported, mas sive; Dmm (s) – diamicton, ma trix sup ported, mas sive, sheared; Sd – sand, de formed; Sl – sand with low-an gle in clined strat i fi ca tion; Sh – sand, hor i zon tally bed ded; Sp –sand, pla nar cross-bed ded; SGp –sandy gravel, pla nar cross-bed ded; Fv – varved clay (lithofacies codes adapted from Miall, 1978 and Eyles et al., 1983)
PURITY OF THE QUARTZ EXTRACTS
In a first test, quartz ex tracts were pre pared as large (8 mm) aliquots and two aliquots per sam ple were stim u lated with in fra - red light to iden tify pos si ble feld spar con tam i na tion. It is con sid - ered that the sen si tiv ity to in fra red stim u la tion is sig nif i cant when the OSL IR de ple tion ra tio is more than 10% of the blue sig nal (Duller, 2003), which was the case for all of our sam ples.
In or der to ob tain a purer quartz sig nal, a post-IR blue Sin gle Aliquot Re gen er a tive (SAR) pro to col (Banerjee et al., 2001) was ap plied. Be tween 24 and 29 aliquots were mea sured for each sam ple to ob tain the equiv a lent dose (De).
DOSE RECOVERY PREHEAT PLATEAU AND PREHEAT PLATEAU TESTS
Three aliquots per tem per a ture from one rep re sen ta tive sam ple (Lªèi 1) were used to cal cu late the dose re cov ery ra tio at dif fer ent pre heat tem per a tures. In that way, eigh teen aliquots cov ered the pre heat range of 180 to 280°C with an in ter val of 20°C. Cut heat was kept at 20°C lower than the pre heat. Only the pre heat tem per a tures at which the dose re cov ery ra tio were within 10% of unity are con sid ered ac cept able. The re sults show that pre heat at 220°C seems to be the most suit able for the Lªèi 1 sam ple (Fig. 3A). To com ple ment the re sults of dose-re cov ery pre heat pla teau, which could be mis lead ing (Rob erts, 2006), we also car ried out a stan dard pre heat pla teau test. Fol low ing this, three aliquots of the same sam ple and with the same tem per a ture set tings were used. From this test, tem - per a ture ranges from 200 to 240°C show a pla teau and the De
value at 220°C re veals the small est un cer tainty. Com bin ing the two tests, the pre heat tem per a ture of 220°C should be used in fur ther anal y ses.
DOSE RECOVERY TEST
To test whether the mea sure ment pro to col was suit able for the other sam ples, a dose re cov ery test was con ducted on all sam ples. Six aliquots per sam ple were bleached in day light on a window sill for a few days, and then ad di tion ally stim u lated us - ing blue LED ex po sure for 60 s to empty the pre vi ous OSL sig - nal. The av er age mea sured to given dose ra tio was 0.94 ± 0.02, 1.04 ± 0.10 and 1.09 ± 0.06 for sam ples Lªèi 1, 2 and Nereta, re spec tively (Ta ble 1), thus im ply ing that the pro to col is able to mea sure ac cu rately a dose given be fore ther mal pre-treat ment (Wintle and Murray, 2006). A higher ra tio of 1.13 ± 0.10 was ob - tained for the Kalna Bñbli sam ple in di cat ing an un suit able pre - heat tem per a ture. Af ter an ad di tional stan dard pre heat pla teau test for that sam ple based on six aliquots (Fig. 3B), we adopted a pre heat tem per a ture of 200°C and a cut heat of 180°C. The dose re cov ery test for that tem per a ture gave an ac cept able ra - tio of 0.94 ± 0.060 (n = 6).
DATA ANALYSIS
RisÝ An a lyst 4.10 soft ware was used to cal cu late equiv a - lent doses (De); only aliquots with test dose er ror <10%, a re cy - cling ra tio within 10% of unity, and a sig nal more than three sigma above the back ground were ac cepted. Re cy cled points were used for ex po nen tial-lin ear fit ting and the growth curve was forced through the or i gin. The fast-com po nent-dom i nated OSL sig nal was iso lated from the first part of the de cay curve, so that the time in ter vals of 0.08–0.32 s and 0.48–1.08 s were used for peak and back ground sig nals, re spec tively. By us ing
an early back ground sub trac tion, any in flu ence of me dium and slow sig nal com po nents was min i mized (Ballarini et al., 2007).
The overdispersion (OD) was cal cu lated us ing an ex cel macro by S. Huot based on the Cen tral Age Model (Galbraith et al., 1999). Fi nally, we used the three-pa ram e ter min i - mum-age model (MAM-3; Galbraith et al., 1999) fol low ing the first cri te rion of the sin gle-aliquot de ci sion pro to col of Ar nold et al. (2007), and this was ful filled by sam ples from Nereta and Kalna Bñbli. For the two Lªèi sam ples the Cen tral Age Model (CAM) was ap plied.
RESULTS
WATER CONTENT AND DOSE RATE
The nat u ral and sat u rated wa ter con tent ranges be tween 0.2–3.5% and 24.4–27.9%, re spec tively (Ta ble 1). All these val - ues seem to be rather un likely as av er age wa ter con tent since de po si tion. Field mea sure ment of the wa ter con tent re veals, in 484 Kristaps Lamsters, Edyta Kaliñska-Nartiša, Vitªlijs Zelès and Helena Alexanderson
Fig. 3. Dose re cov ery ra tio at dif fer ent pre heat tem per a tures (in black) and pre heat pla teau test (in red) for (A) the Lªèi 1 (14051) sam ple, and (B) the Kalna Bñbli (14054) sam ple ex clud ing a tem per a ture of 220°C
con trast, val ues be tween 3.4 and 8.5% (Kalna Bñbli and Lªèi 2, re spec tively). We, there fore, es ti mated a life time av er age burial wa ter con tent of 9 ± 4% for all sam ples, as sum ing that sat u - rated wa ter con di tions oc curred for ap prox i mately less than half of the burial time. The un cer tainty takes into ac count in ev i ta ble fluc tu a tions in wa ter con tent with time. A nar row range of dose rates be tween 1.36 ± 0.07 Gy/ka and 1.6 ± 0.08 Gy/ka (for Lªèi 1 and Nereta, re spec tively) was ob tained.
LUMINESCENCE SIGNAL CHARACTERISTICS
The lu mi nes cence sig nal is dom i nated by the fast OSL com po nent that can be de tected from the fast de cay (cf. de cay curve for Lªèi 1 sam ple, Fig. 4). How ever, me dium-to-slow com po nents also oc cur in some aliquots and this can be de - tected in the Nereta sam ple (Fig. 4). Al though two growth curves re veal an ac cept able re cy cling and low re cu per a tion (Fig. 4), oc ca sion ally the lim its of both re cy cling and re cu per a - tion were ex ceeded. The av er age re cy cling ra tio is 1.01 ± 0.004 (n = 102) and the av er age re cu per a tion is 0.8 ± 0.2 (n = 102).
DOSE DISTRIBUTION AND AGE CALCULATION
Equiv a lent dose (De) dis tri bu tions for in di vid ual sam ples are broad; the dif fer ence be tween the low est and the high est aliquot De is 90–290 Gy (Fig. 5). The mean De is 152 ± 13 Gy, 131 ± 11, 90 ± 8 Gy and 144 ± 13 Gy (Lªèi 1, Lªèi 2, Nereta and Kalna Bñbli, re spec tively). Whereas the Lªèi 1 dose dis tri bu tion is sym met ri cal, for the other sam ples the dis tri bu tions are ei ther slightly neg a tively skewed (Sk = -0.2 at Lªèi 2) or sig nif i cantly pos i tively skewed (Sk = 1.4 and 1.1 at Nereta and Kalna Bñbli, re spec tively; Fig. 5). For all sam ples, a high overdispersion (OD) was ob tained (42–46%; Ta ble 2). Both the cen tral (CAM) and min i mum-age model (MAM-3) doses stay in good agree - ment with mean De, al though be ing slightly youn ger.
The four fi nal OSL ages (Ta ble 2) ob tained from the LLL fall into three groups that vary be tween 56 ± 6 ka and 112 ± 11 ka (Nereta and Lªèi 1), with two re sults around 90 ka (92 ± 9 ka for Lªèi 2 and 93 ± 10 ka for Kalna Bñbli). The CAM and MAM-3 ages, in con trast, show ranges of 82 ± 10 ka and 102 ± 12 ka (Lªèi 1 and 2), and 44 ± 10 ka and 84 ± 9 ka (Nereta and Kalna Bñbli).
INTERPRETATION
The rou tine tests of as sump tion i.e. dose re cov ery test (cf.
Re sults; Ta ble 1) re veal that the SAR pro to col can suc cess fully mea sure a given lab o ra tory dose. We are thus con fi dent that
Field ID LLL ID
238U ± s.e.
[Bq/kg]
226Ra ± s.e.
[Bq/kg]
232Th ± s.e.
[Bq/kg]
40K ± s.e.
[Bq/kg] w.c. field/
nat u ral [%] w.c. sat.
[%] w.c.
[%]
To tal dose rate ± s.e.
[Gy/ka]
Dose re cov ery Lªèi 1 14051 –3.36 ± 8.71 7.36 ± 0.63 7.25 ± 0.59 366 ± 14 3.5/3.5 27.9 9 ± 4 1.36 ± 0.07 0.94 ± 0.02 Lªèi 2 14052 4.93 ± 5.28 9.35 ± 0.39 11.52 ± 0.37 361 ± 8 8.5/1.9 24.4 9 ± 4 1.42 ± 0.07 1.04 ± 0.10 Nereta 14053 –1.70 ± 8.00 10.27 ± 0.61 11.47 ± 0.57 421 ± 14 4.2./0.2 25.6 9 ± 4 1.60 ± 0.08 1.09 ± 0.06 Kalna
Bñbli 14054 –4.04 ± 7.57 9.91 ± 0.57 12.65 ± 0.53 395 ± 12 3.4/1.3 24.6 9 ± 4 1.54 ± 0.08 0.94 ± 0.06
s.e. – stan dard er ror; w.c. – wa ter content
T a b l e 1 Sum mary of radionuclide con cen tra tions, wa ter con tents, to tal dose rates and dose re cov ery ra tios
Fig. 4. Examples of growth and decay curves for the Lªèi 1 (14051) and Nereta (14053) samples
the good ac cu racy of each De’s is pro duced by the SAR pro to - col. How ever, a few po ten tial is sues need to be ad dressed.
Wa ter and gla cier-trans ported sed i ments, due to their depositional en vi ron ment, tend to suf fer from in com plete bleach ing (cf. Rodnight et al., 2006; Fuchs and Owen, 2008;
Alexanderson and H¯kansson, 2014; King et al., 2014; Mehta et al., 2014). A sin gle-aliquot ap proach re ported ev i dence of an
asym met ri cal De dis tri bu tion (cf. Re sults) that is likely due to the pres ence of poorly bleached grains in the sam ple (Thrasher et al., 2009). Be cause the sam ples from the Nereta and Kalna Bñbli quar ries have clearly asym met ri cal and largely pos i tively skewed De dis tri bu tions (Fig. 5), poorly bleached quartz grains seem to con trib ute to the dose here. Con sid er ing, how ever, the old OSL ages of the de pos its, the ef fect of in com plete bleach ing 486 Kristaps Lamsters, Edyta Kaliñska-Nartiša, Vitªlijs Zelès and Helena Alexanderson
Fig. 5. Dose distribution and density probability function of the samples investigated
Field ID Lab ID* n De ± s.e.
[Gy] Age ± s.e.
[ka] Depth of sam ple
De ± s.e.
(CAM ** or MAM-3; Gy)
Age ± s.e.
(CAM** or MAM-3; ka) p
Lªèi 1 Lund-14051 24 152 ± 13 112 ± 11 4.7 138 ± 14** 102 ± 12** –
Lªèi 2 Lund-14052 24 131 ± 11 92 ± 9 9.2 117 ± 12** 82 ± 10** –
Nereta Lund-14053 29 90 ± 8 56 ± 6 5.5 70 ± 16 44 ± 10 0.9
Kalna
Bñbli Lund-14054 25 144 ± 13 93 ± 10 5.0 130 ± 13 84 ± 9 0.9
* – Lund Lu mi nes cence Lab o ra tory, Swe den; ** – sam ples in which the CAM was used; n – num ber of aliquots mea sured to ob tain av er age De’s; s.e. – stan dard er ror; p – prob a bil ity val ues; the most re li able ages are in bold
T a b l e 2 Sum mary of equiv a lent doses (De) and ages
is usu ally ex pected to be small (Murray and Olley, 2002). Ad di - tion ally, the ef fect of a few poorly bleached grains, in case where the ma jor ity of the grains are bleached, is rather small (Alexanderson and Murray, 2012).
The overdispersion (OD) value, which re flects the spread of mul ti ple in de pend ent De es ti mates from aliquots (Thomas and Burrough, 2013) can also be used to char ac ter ise the depositional en vi ron ment. For ex am ple, to in di cate fac tors such as in suf fi cient ex po sure to sun light prior to burial (Olley et al., 1996), post-depositional mix ing of grains (Jacobs, 2008), re - cent ero sion from the bed rock (Lukas et al., 2007) or microdosimetry in het er o ge neous and coarse-grained de pos its (Klasen et al., 2007).
In this study, the OD is high com pared to e.g. well-bleached sam ples (Ar nold and Rob erts, 2009), and of a very sim i lar value among the sites (Ta ble 2). This can be ex plained by a sim i lar op por tu nity of sed i ment bleach ing and depositional con di tions and/or source sed i ments. Com bined with the asym met ri cal na - ture of De with large val ues of OD of the de pos its in ves ti gated, a par tial bleach ing of their grains can not be re jected. How ever, these ages are based on large aliquot mea sure ments, and thus even if in com plete bleach ing was pres ent, it is likely to be ob - scured by the ef fects of av er ag ing be tween grains (Duller, 2008). The ages suf fer from rel a tively poor pre ci sion. This is likely due to the wide spread of equiv a lent dose dis tri bu tions re - sult ing in high OD val ues. No data on OD are avail able for sim i - lar sed i men tary units at nearby sites, thus hin der ing a pos si ble cor re la tion. How ever, con sid er ing the wide spa tial dis tri bu tion of the sites in ves ti gated, such a high OD value in all of them is thought to be due to the gen eral lu mi nes cence na ture of the de - pos its in ves ti gated, and lo cal fac tors should be ruled out.
Slightly tighter and more sym met ri cal dose dis tri bu tions were found for the Lªèi 1 and 2 sam ples, thus pos si bly sug gest ing their ho mo ge neous bleach ing (cf. Murray and Olley, 2002).
Con versely, a wide and sym met ri cal De dis tri bu tion may in di - cate a large pro por tion of in com pletely bleached grains (Wallinga, 2002). But whether or not the de posit at the Lªèi site is in com pletely bleached can not be re solved at this stage of re - search. Lim ited re sults from the neigh bour ing Krimñnas site (unpubl. data; Oinonen and Eskola, 2009), how ever, largely co - in cide with the ages at the Lªèi site and, there fore, to some ex - tent ar gue for a well-bleached de posit at Lªèi.
Mostly only fast sig nal com po nents that re set quickly (Kuhns et al., 2000), and are quick to bleach un der op ti cal stim - u la tion (Thrasher et al., 2009), dom i nate among the sam ples in - ves ti gated (Fig. 4). Al though Jain et al. (2003) re ported on the dom i nance of the fast quartz com po nent in most quartz sam - ples, some stud ies, how ever, showed that sed i ments could suf - fer from a lack of the fast com po nent, which re sults in the pres - ence of dif fi cult-to-bleach me dium-to-slow com po nents (Lukas et al., 2007). Some of the aliquots from the Nereta sam ple do show such me dium-to-slow com po nents (Fig. 4) but there is also a fast sig nal com po nent pres ent. In the rest of our sam ples the fast-sig nal com po nent seems to be strong.
The MAM sta tis ti cally se lects the aliquots with the low est De’s, that is, those that were most prob a bly suf fi ciently bleached (Wyshnytzky et al., 2015), and in this study pro vides 9–12 ka youn ger ages (Ta ble 2); this is, how ever, within the er ror of pre - sented mean ages. The prob a bil ity (p) value for the Nereta and Kalna Bñbli sam ples is high and in these cases the MAM has pro duced re li able ages. This is ad di tion ally sup ported by the De’s dis tri bu tion, where the peaks are closer to the MAM value (Fig. 5). We there fore fa vour the MAM ages for the Nereta and Kalna Bñbli. Ages ob tained from the CAM (for the Lªèi 1 and 2) are also youn ger (10 ka) and within er ror.
The wa ter con tent in the de posit in flu ences the dose rate and hence the fi nal age (Wintle, 2008). If as sum ing the pres - ent-day wa ter con tent as mea sured in the field (cf. Re sults) as an av er age wa ter con tent, the age range for the 53–112 ka sam ples would be 53–106 ka, where the Lªèi 2 sam ple would keep the same age. Con versely, con sid er ing the other ex treme – sat u rated con di tions – the range would be come 65–131 ka.
Re sult ing from this, most ages would be within er ror of the pres - ent se lected ages and does not change our main con clu sions of the age of the de pos its. Only the age ob tained from the Lâèi 1 sam ple in sat u rated con di tions is within two stan dard er rors of the orig i nal age of 112 ± 11 ka. The un cer tain ties in wa ter-con - tent es ti mates thus do not change our chro nol ogy sig nif i cantly.
DISCUSSION
Our new OSL ages com ple ment the Early and Mid dle Weichselian his tory of cen tral Lat via by pro vid ing the time when ice-free con di tions pre vailed. The OSL ages from the Lªèi and Kalna Bñbli sites range be tween 84 ± 9 ka and 112 ± 11 ka and fall within ma rine iso tope stage 5 (MIS 5). This cor re sponds to the Early Weichselian (117–75 ka af ter Mangerud, 1991, or 115–85 ka af ter Svendsen et al., 2004) in the North Eu ro pean chronostratigraphy.
Our find ings in Lat via thus sup port data from Es to nia and Lith u a nia by show ing that dur ing Early Weichselian time the Bal tic ter ri tory ex pe ri enced ice-free con di tions at least from 115 ka to 68 ka (Kalm et al., 2011) or from 114.3 ± 7.4 to 76.5 ± 4.9 ka (Molodkov et al., 2010). The lu mi nes cence re cord from the Early Weichselian in Lat via is oth er wise so far rather scarce, but the ex ist ing data is in agree ment with our re sults.
Lac us trine de pos its over lain by Weichselian till in SE Lat via (¦idiÖi) have been dated by thermoluminescence (TL) to 79 ka (Meirons et al., 1981). Lac us trine and flu vial de pos its from Subate and Zvidziena (see Fig. 1 for lo ca tion) re vealed an age range be tween 92 and 97 ka (Meirons, 1986). Of Early and Mid - dle Weichselian age are also the glaciofluvial de pos its at LejaslabiÖi and CÂre in north west ern Lat via (Zelès et al., 2011), which yielded four lu mi nes cence ages of 85 ± 5, 42 ± 3, 115 ± 10 and 45 ± 3 ka. Fi nally, at the Krimñnas site, lo cated some 14 km north of the Lªèi site, sed i ments were de pos ited at a sim i lar time frame (101 ± 18 ka and 95 ± 15 ka; unpubl. data;
Oinonen and Eskola, 2009).
The OSL age from the Nereta site (44 ± 10 ka), on the other hand, cor re sponds with the Mid dle Weichselian (MIS 3). Dur ing the Mid dle Weichselian a sug gested gla ci ation in Lat via known as the Talsi stadial (Zelès and Markots, 2004) could have reached west ern and north ern Lat via be tween 68 ka and 54 ka (Zelès et al., 2011). It is most likely that this gla cial event did not ex tend into the mid dle and south ern part of the study area ac - cord ing to lo cal (Zelès et al., 2011) and re gional re con struc tions (Svendsen et al., 2004). Ac cord ing to mod el ling of the Weichselian Ice Sheet ad vances in the Bal tic re gion by Holmlund and Fastook (1995), the first gla cial ad vance into the Gulf of Riga was prob a bly at about 64 ka. It eroded the up per part of the Eemian and Early Weichselian ma rine clay con tain - ing Portlandia arctica shells and as sim i lated these into the low - er most Weichselian till unit (Molodkov et al., 1998), which is ex - posed along the bluffs of the Daugava River Val ley in the vi cin - ity of Daugmale, north ern CLL (Dreimanis and Zelès, 1998, 2004).
In com bi na tion with the lack of ev i dence of youn ger gla cial de po si tion in the mid dle and south ern plains of CLL, it seems that no gla ci ation took place in this part of Lat via be tween the
Eemian Inter gla cial and the Late Weichselian. Be tween 54 ka and 24 ka ice-free con di tions per sisted through out all Lat via;
this Mid dle Weichselian interstadial is lo cally named the Lejasciems interstadial (Zelès and Markots, 2004). It is also sug gested from Lith u a nian stud ies that most likely the ma jor part of the East ern Bal tic was not cov ered by ice un til the Late Weichselian (GuobytÅ and Satkñnas, 2011). The ice-free pe - riod in Es to nia was from at least 115 to 68 ka dur ing the Early Weichselian and be tween 44 and 27 ka dur ing the Mid dle Weichselian (Kalm et al., 2011).
Dur ing the long in ter val from the Early Weichselian to the be gin ning of the Late Weichselian, fa vour able con di tions could have ex isted in cen tral Lat via for the de po si tion of shal low ba sin and flu vial sed i ments, which have been dated in sev eral lo cal i - ties else where in Lat via and Lith u a nia (Satkñnas et al., 2009, 2013; Zelès et al., 2011; Saks et al., 2012). The OSL-dated de - pos its in our study could re flect a sim i lar en vi ron ment, how ever, most of the ages pre sented in this study, with the ex cep tion of the Nereta sam ple, do not co in cide with those of the other de - pos its (e.g., Saks et al., 2012) of Mid dle Weichselian age. This could be ex plained by the pos si ble ero sion of these youn ger sed i ments in the cen tral part of the Zemgale Ice Lobe, which ad vanced sev eral times in the Late Weichselian. Gla cial ero - sion was less ef fec tive else where; sed i ments de pos ited dur ing the Mid dle Weichselian are known from other lo cal i ties in Lat - via, where gla cial ero sion was weaker. For ex am ple, OSL dates (Zelès et al., 2011) from sandy de pos its found in an interlobate po si tion at the Kažoki site (see Fig. 1 for lo ca tion) and lo cated on the east ern slope of the East ern Kursa Up land, ad ja cent to the study area, yield ages of 26.9 ± 4.4 ka and 29.4 ± 4.7 ka.
These youn ger, ex ten sively folded and thrusted outwash de - pos its were pre served dur ing the ad vance of the Zemgale Ice Lobe; this is most likely due to weaker gla cial ero sion on the pe - riph ery of the East ern Kursa Up land.
Fi nally, ESR dates from Portlandia arctica shells from till in the banks of the Daugava River, Daugmales TomÂni site, pro - vide ad di tional strati graphi cal ev i dence for Pleis to cene de po si - tion in the CLL (Molodkov et al., 1998). The ESR ages
(86.0 ± 6.8 ka, 105.0 ± 9.2 ka) sup port the clear pat tern of ice-free con di tions dur ing the Early Weichselian as noted in this study. This is be cause P. arctica shells come from Early Weichselian interstadial ma rine de pos its that were dis placed from their or i gin in the Gulf of Riga by the Zemgale Lobe dur ing the last gla ci ation (Molodkov et al., 1998, Zelès et al., 2011).
CONCLUSIONS
Our study dem on strates the ex tent to which lu mi nes cence dat ing can con trib ute to deeper un der stand ing of the his tory of the largely un stud ied sandy de pos its over lain by subglacial till of the Late Weichselian in cen tral Lat via. Three OSL ages from the area cov ered by the Zemgale Ice Lobe range be tween 84 ± 9 ka and 112 ± 11 ka, cor re spond ing to an Early Weichselian age (MIS 5), and are sup ported by un pub lished OSL ages from nearby sites (un pub lished data; Oinonen and Eskola, 2009). The OSL dat ing re sults ob tained dem on strate ice-free con di tions dur ing the Early and Mid dle Weichselian, when de po si tion of sandy sed i ments oc curred in cen tral Lat via.
We found no Mid dle Weichselian age de pos its in the Cen tral Lat vian Low land and sup pose that they could pos si bly have been eroded dur ing the ad vances of the Zemgale Ice Lobe in the Late Weichselian, as this area was sub ject to con sid er able gla cial ero sion. How ever, Mid dle Weichselian (MIS 3) de pos its (one sam ple; 44 ± 10 ka) were found in the Nereta Quarry, which is lo cated in the glaciotectonised ice-mar ginal suc ces sion next to the interlobate area.
Ac knowl edge ments. The re search was par tially sup - ported by the Swed ish In sti tute, pro ject 00365.2014 to E. Kaliñska-Nartiša, and by the Uni ver sity of Lat via pro ject “Cli - mate change and sus tain able use of nat u ral re sources” (No.
AAP2016/B041) to K. Lamsters. Dr. E. Thamó-Bozsó (Geo log i - cal and Geo phys i cal In sti tute of Hun gary) and an anon y mous re viewer are thanked for valu able com ments, which im proved the manu script.
REFERENCES
Alexanderson, H., H¯kansson, L., 2014. Coastal gla ciers ad - vanced onto Jameson Land, East Green land dur ing the late gla - cial early Ho lo cene Milne Land Stade. Po lar Re search, 33:
1–17.
Alexanderson, H., Murray, A.S., 2012. Prob lems and po ten tial of OSL dat ing Weichselian and Ho lo cene sed i ments in Swe den.
Qua ter nary Sci ence Re views, 44: 37–50.
Ar nold, L.J., Rob erts, R.G., 2009. Sto chas tic mod el ling of multi-grain equiv a lent dose (De) dis tri bu tions: im pli ca tions for OSL dat ing of sed i ment mix tures. Qua ter nary Geo chron ol ogy, 4: 204–230.
Ar nold, L.J., Bailey, R.M., Tucker, G.E., 2007. Sta tis ti cal treat ment of flu vial dose dis tri bu tions from south ern Col o rado ar royo de - pos its. Qua ter nary Geo chron ol ogy, 2: 162–167.
boltiÖë, O., 1963. Gorizonty verkhney moreny Zemgalyskoy ravniny i vopros o formirovanii Linkuvskoy konechnoy moreny (in Rus sian). In: Ucheniye zapiski aspirantov Latviyskogo Gosudarstvennogo Universiteta im. P. Stuchki (ed. N.S.
Temnikova), 1: 5–17.
boltiÖë, O., 1989. Glaciostruktura i lednikoviy morfogenez (in Rus sian). Zinªtne, Rîga.
boltiÖë, O., Vaitekunas, P., Danilªns, I., Karukapp, R., Klive, G., Raukas, A., Roshko, L., Chebotareva, N.S., Yanke, V., 1977.
The Bal tic ice stream. In: The struc ture and dy nam ics of the last ice sheet of Eu rope (ed. N.S. Chebotareva): 17–44. Nauka, Mos cow.
Ballarini, M., Wallinga, J., Wintle, A.G., Bos, A.J.J., 2007. A mod i - fied SAR pro to col for op ti cal dat ing of in di vid ual grains from young quartz sam ples. Ra di a tion Mea sure ments, 42: 360–369.
Baltrñnas, V., Karmaza, B., Dundulis, K., Gadeikis, S., Raèkauskas, V., inkñnas, P., 2005. Char ac ter is tic of till for - ma tion dur ing Baltija (Pom er a nian) stage of the Nemunas (Weichselian) gla ci ation in Lith u a nia. Geo log i cal Quar terly, 49 (4): 417–428.
Baltrñnas, V., eirienÅ, V., Molodkov, A., ZinkutÅ, R., Katinas, V., Karmaza, B., KisielienÅ, D., Petroëius, R., Taraëkevièius, R., Pilièiauskas, G., Schmölcke, U., Hein rich, D., 2013.
Depositional en vi ron ment and cli mate changes dur ing the late Pleis to cene as re corded by the Netiesos sec tion in south ern Lith u a nia. Qua ter nary In ter na tional, 292: 136–149.
Banerjee, D., Murray, A. S., BÝtter-Jensen, L., Lang, A. 2001.
Equiv a lent dose es ti ma tion us ing a sin gle aliquot of polymineral fine grains. Ra di a tion Mea sure ments, 33: 73–94.
488 Kristaps Lamsters, Edyta Kaliñska-Nartiša, Vitªlijs Zelès and Helena Alexanderson
BÝtter-Jensen, L., Bulur, E., Duller, G.A.T., Murray, A.S., 2000.
Ad vances in lu mi nes cence in stru ment sys tems. Ra di a tion Mea - sure ments, 32: 523–528.
Brangulis, A.J., Kurës, V., Misªns, J., Stinkulis, G., 1998. Geo - log i cal Map of Lat via, Scale 1:500,000. Geo log i cal Struc ture De - scrip tion (in Lat vian). State Geo log i cal Sur vey, RÌga.
Danilªns, I., 1973. Chetvertichnye otlozheniya Latvii (in Rus sian).
Zinªtne, RÌga.
Dreimanis, A., 1935. The rock de for ma tions, caused by in land ice, on the left bank of Daugava at Dole Is land, Near RÌga in Lat via (in Lat vian with Eng lish sum mary). Gulbis, RÌga.
Dreimanis, A., 1992. Down ward in jected till wedges and up ward in - jected till dikes. Sveriges Geologiska Undersökning, 4: 91–96.
Dreimanis, A., Zelès, V., 1995. Pleis to cene stra tig ra phy of Lat via.
In: Gla cial De pos its in North-East Eu rope (eds. J. Ehlers, S.
Kozarski and P.L. Gibbard): 105–113. Balkema, Rot ter - dam/Brookfield.
Dreimanis, A., Zelès, V., 1998. Daugmale ribbed mo raine: in tro duc - tion to STOP 1. Stop 1: In ter nal struc ture and mor phol ogy of glaciotectonic land forms at Daugmale. In: The INQUA Peribaltic Group Field Sym po sium on Gla cial Pro cesses and Qua ter nary En vi ron ment in Lat via, May 25–31, 1998, RÌga, Lat via. Ex cur - sion guide (ed. V. Zelès): 3–14. Uni ver sity of Lat via, RÌga.
Dreimanis, A., Zelès, V., 2004. STOP 13: The River Daugava Bluff Sec tions at the Ridge Kalnmuiža and Hill Kranèkalns. In: In ter - na tional Field Sym po sium on Qua ter nary Ge ol ogy and Mod ern Ter res trial Pro cesses, West ern Lat via, Sep tem ber 12–7, 2004:
Ex cur sion Guide (ed. V. Zelès): 70–75. Uni ver sity of Lat via, RÌga.
Duller, G.A.T., 2003. Dis tin guish ing quartz and feld spar in sin gle grain lu mi nes cence mea sure ments. Ra di a tion Mea sure ments, 37: 161–165.
Duller, G.A.T., 2008. Sin gle-grain op ti cal dat ing of Qua ter nary sed i - ments: why aliquot size mat ters in lu mi nes cence dat ing. Boreas, 37: 589–612.
Eyles, N., Eyles, C.H., Miall, A.D., 1983. Lithofacies types and ver - ti cal pro file mod els: an al ter na tive ap proach to the de scrip tion and en vi ron men tal in ter pre ta tion of gla cial diamict and diamictite se quences. Sedimentology, 30: 393–410.
Fuchs, M., Owen, L.A., 2008. Lu mi nes cence dat ing of gla cial and as so ci ated sed i ments: re view, rec om men da tions and fu ture di - rec tions. Boreas, 37: 636–659.
Gaigalas, A., 1995. Gla cial his tory of Lith u a nia. In: Gla cial De pos its in North-East Eu rope (eds. J. Ehlers, S. Kozarski and P.L.
Gibbard): 127–137. Balkema, Rot ter dam.
Gaigalas, A., Marcinkevièius, V., 1982. Bed ded struc ture and gen - e sis of forms of hol low-ridge gla cial re lief in North Lith u a nia.
Geologija, 3: 69–79.
Galbraith, R.F., Rob erts, R.G., Laslett, G.M., Yoshida, H., Olley, J.M., 1999. Op ti cal dat ing of sin gle and mul ti ple grains of quartz from Jinmium rock shel ter, North ern Aus tra lia: Part I, Experimetal de sign and sta tis ti cal mod els. Archaeometry, 41:
339–364.
Ginters, G., 1978. Moreny Yuzhno-Kurzemskoy nizmennosti. In:
Problemy morfogeneza relyefa i paleogeografii Latvii (eds. O.
boltiÖë, V. Klªne and G. Eberhards) (in Rus sian): 99–107. Lat - vian State Uni ver sity im. P. Stuchki, RÌga.
GuobytÅ, R., 2004. A brief out line of the Qua ter nary of Lith u a nia and the his tory of its in ves ti ga tion. In: Ex tent and Chro nol ogy of Glaciations, 1 (Eu rope) (eds. J. Ehlers and P.L. Gibbard):
245–250. Elsevier, Am ster dam.
GuobytÅ, R., Satkñnas, J., 2011. Pleis to cene glaciations in Lith u a - nia. In: Qua ter nary Glaciations – Ex tent and Chro nol ogy: A Closer Look, 15 (eds. J. Ehlers, P.L. Gibbard and P.D. Hughes):
231–246. Elsevier, Am ster dam.
Holmlund, P., Fastook, J.L., 1995. A time de pend ent glaciological model of the Weichselian ice sheet. Qua ter nary In ter na tional, 27: 53–58.
Jacobs, Z., 2008. Lu mi nes cence chro nol o gies for coastal and ma - rine sed i ments. Boreas, 37: 508–535.
Jain, M., BÝtter-Jensen, L., Singhvi, A.K., 2003. Dose eval u a tion us ing mul ti ple-aliquot quartz OSL: Test of meth ods and a new pro to col for im proved ac cu racy and pre ci sion. Ra di a tion Mea - sure ments, 37: 67–80.
Juškeviès, V., 2000. Qua ter nary Geo log i cal Map (in Lat vian with Eng lish sum mary). In: Geo log i cal map of Lat via, scale 1:200,000. Sheet 43 – RÌga, Sheet 53 – Ainaûi. Ex plan a tory Notes and Maps (eds. O. boltiÖë and V. Kurës). State Geo log i - cal Sur vey, RÌga.
Juškeviès, V., 2001. Qua ter nary Geo log i cal Map (in Lat vian with Eng lish sum mary). In: Geo log i cal map of Lat via, scale 1:200,000. Sheet 32 – Jelgava. Ex plan a tory Notes and Maps (eds. O. boltiÖë and A. Brangulis). State Geo log i cal Sur vey, RÌga.
Kaliñska-Nartiša, E., Nartišs, M., Thiel, C., Buylaert, J.-P., Murray, A.S., 2015a. Late-gla cial to Ho lo cene ae olian de po si - tion in north east ern Eu rope – The tim ing of sed i men ta tion at the Iisaku site (NE Es to nia). Qua ter nary In ter na tional, 357: 70–81.
Kaliñska-Nartiša, E., Thiel, C., Nartišs, M., Buylaert, J.-P., Murray, A.S., 2015b. Age and sed i men tary re cord of in land ae - olian sed i ments in Lith u a nia, NE Eu ro pean Sand Belt. Qua ter - nary Re search, 84: 82–95.
Kaliñska-Nartiša, E., Thiel, C., Nartišs, M., Buylaert, J.-P., Murray, A.S., 2016. The north-east ern ae olian “Eu ro pean Sand Belt” as po ten tial re cord of en vi ron men tal changes: a case study from East ern Lat via and South ern Es to nia. Ae olian Re search, 22: 59–72.
Kalm, V., 2006. Pleis to cene chronostratigraphy in Es to nia, south - east ern sec tor of the Scan di na vian gla ci ation. Qua ter nary Sci - ence Re views, 25: 960–975.
Kalm, V., 2012. Ice-flow pat tern and ex tent of the last Scan di na vian Ice Sheet south east of the Bal tic Sea. Qua ter nary Sci ence Re - views, 44: 51–59.
Kalm, V., Raukas, A., Rattas, M., Lasberg, K., 2011. Pleis to cene glaciations in Es to nia. De vel op ments in Qua ter nary Sci ence, 15: 95–104.
King, G.E., Rob in son, R.A.J., Finch, A.A., 2014. To wards suc - cess ful OSL sam pling strat e gies in gla cial en vi ron ments: de ci - pher ing the in flu ence of depositional pro cesses on bleach ing of mod ern gla cial sed i ments from Jostedalen, South ern Nor way.
Qua ter nary Sci ence Re views, 89: 94–107.
Klasen, N., Fiebig, M., Preusser, F., Reitner, J.M., Radtke, U., 2007. Lu mi nes cence dat ing of proglacial sed i ments from the East ern Alps. Qua ter nary In ter na tional, 164–165: 21–32.
Kuhns, C.K., Agersnap Larsen, N., McKeever, S.W.S., 2000.
Char ac ter is tics of LM-OSL from sev eral dif fer ent types of quartz. Ra di a tion Mea sure ments, 32: 413–418.
Lamsters, K., 2012. Drum lins and re lated glaciogenic land forms of the Madliena Tilted Plain, Cen tral Lat vian Low land. Bul le tin of the Geo log i cal So ci ety of Fin land, 84: 45–57.
Lamsters, K., Zelès V., 2015. Subglacial bedforms of the Zemgale Ice Lobe, SE Bal tic. Qua ter nary In ter na tional, 386: 42–54.
Lasberg, K., Kalm, V., 2013. Chro nol ogy of Late Weichselian gla ci - ation in the west ern part of the East Eu ro pean Plain. Boreas, 42:
995–1007.
Levkov, E., 1980. Glaciotektonika (in Rus sian). Nauka i tekhnika, Minsk.
Lukas, S., Spencer, J.Q.G., Rob in son, R.A.J., Benn, D.I., 2007.
Prob lems as so ci ated with lu mi nes cence dat ing of Late Qua ter - nary gla cial sed i ments in the NW Scot tish High lands. Qua ter - nary Geo chron ol ogy, 2: 243–248.
Mangerud, J., 1991. The last inter gla cial/gla cial cy cle in north ern Eu rope. In: Qua ter nary Land scapes (eds. L.C.K. Shane and E.J. Cush ing): 38–75. Uni ver sity of Min ne sota Press, Min ne ap - o lis.
Mehta, M., Dobhal, D., Pratap, B., Majeed, Z., Gupta, A. K., Srivastava, P., 2014. Late Qua ter nary gla cial ad vances in the Tons River Val ley, Garhwal Himalaya, In dia and re gional synchronicity. The Ho lo cene, 24: 1336–1350.
Meirons, Z., 1986. Stra tig ra phy of Pleis to cene de pos its in Lat via (in Rus sian). In: Issledovaniya lednikovykh obrazovaniy Pribaltiki (eds. O. Kondratiene and A. Mikalauskas): 69–81. AN Litovskoi SSR, Vilnius.
Meirons, Z., 2002. Qua ter nary Geo log i cal Map. In: Geo log i cal map of Lat via, scale 1:200,000. Sheet 33 – Ogre. Ex plan a tory Notes and Maps (eds. O. boltiÖë and A. Brangulis) (in Lat vian with Eng lish sum mary). State Geo log i cal Sur vey, RÌga.
Meirons, Z., Pun ning, M.J., Hütt, G., 1981. Re sults ob tained through the TL dat ing of South-East Lat vian Pleis to cene de pos - its (in Rus sian with Eng lish sum mary). Eesti NSV Teaduste Akadeemia Toimetised, Geologia 30: 28–33.
Miall, A.D., 1978. Lithofacies types and ver ti cal pro file mod els in braided river de pos its: a sum mary. Ca na dian So ci ety of Pe tro - leum Ge ol o gists Mem oir, 5: 597–604.
Mironovs, G., Vªcele, V., Vasiljeva, V., Karpickis, V., KipÂna, M., 1962. Geologicheskoye stroyeniye i gidrogeologicheskoye usloviya teritorii lista O–35–XXVI (in Rus sian). Otchot Vidzemskoy kompleksnoy GSP po rabotam, 1959–1960 g., I.
Lat vian Geo log i cal Board, RÌga.
Molodkov, A., Dreimanis, A., boltiÖë, O., Raukas, A., 1998. The age of Portlandia arctica shells from gla cial de pos its of Cen tral Lat via: an an swer to a con tro versy on the age and gen e sis of their en clos ing sed i ments. Qua ter nary Geo chron ol ogy, 17:
1077–1094.
Molodkov, A., Bitinas, A., DamuëytÅ, A., 2010. IR-OSL stud ies of till and inter-till de pos its from the Lith u a nian Mar i time Re gion.
Qua ter nary Geo chron ol ogy, 5: 263–268.
Murray, A.S., Olley, J.M., 2002. Pre ci sion and ac cu racy in the op ti - cally stim u lated lu mi nes cence dat ing of sed i men tary quartz: a sta tus re view. Geochronometria, 21: 1–16.
Murray, A.S., Wintle, A.G., 2000. Lu mi nes cence dat ing of quartz us ing an im proved sin gle-aliquot re gen er a tive-dose pro to col.
Ra di a tion Mea sure ments, 32: 57–73.
Murray, A.S., Mar ten, R., Johnston, A., Mar tin, P., 1987. Anal y sis for nat u rally oc cur ring radionuclides at en vi ron men tal con cen - tra tions by gamma spec trom e try. Jour nal of Radioanalytical and Nu clear Chem is try, 115: 263–288.
Oinonen, M., Eskola, K.O., 2009. Re search re port – Lu mi nes cence dat ing of Lat vian sam ples, Pur chase iden ti fi ca tion No. LU 2008/I-269, part 1 (16 sam ples), Hel sinki, p. 5.
Olley, J.M., Murray, A.S., Rob erts, R.G., 1996. The ef fects of disequilibria in the ura nium and tho rium de cay chains on burial dose rates in flu vial sed i ments. Qua ter nary Sci ence Re views, 15: 751–760.
Rattas, M., Kalm, V., Kihno, K., Liivrand, E., Tinn, O., Tänavsuu-Milkeviciene, K., Sakson, M., 2010. Chro nol ogy of Late Saalian and Mid dle Weichselian ep i sodes of ice-free lac us - trine sed i men ta tion re corded in the Arumetsa sec tion, south - west ern Es to nia. Es to nian Jour nal of Earth Sci ences, 59:
125–140.
Raukas, A., Stankowski, W.T.J., Zelès, V., Šinkunas, P., 2010.
Chro nol ogy of the last deglaciation in the South-East ern Bal tic Re gion on the ba sis of re cent OSL dates. Geochronometria, 36:
47–54.
Rinterknecht, V.R., Clark, P.U., Raisbeck, G.M., Yiou, F., Bitinas, A., Brook, E J., Marks, L., Zelès, V., Lunkka, J.-P., Pavlovskaya, I.E., Piotrowski, J.A., Raukas, A., 2006. The last deglaciation of the south east ern sec tor of the Scan di na vian Ice Sheet. Sci ence, 311: 1449–1452.
Rinterknecht, V.R., Bitinas, A., Clark, P.U., Raisbeck, G.M., Yiou, F., Brook, E.J., 2008. Tim ing of the last deglaciation in Lith u a - nia. Boreas, 37: 426–433.
Rob erts, H.M., 2006. Op ti cal dat ing of coarse-silt sized quartz from loess: Eval u a tion of equiv a lent dose de ter mi na tions and SAR pro ce dural checks. Ra di a tion Mea sure ments, 41: 923–929.
Rodnight, H., Duller, G.A.T., Wintle, A.G., Tooth, S., 2006. As - sess ing the reproducibility and ac cu racy of op ti cal dat ing of fluvial de pos its. Qua ter nary Geo chron ol ogy, 1: 109–120.
Saks, T., Kalvªns, A., Zelès, V., 2012. OSL dat ing of Mid dle Weichselian age shal low ba sin sed i ments in West ern Lat via, East ern Bal tic. Qua ter nary Sci ence Re views, 44: 60–68.
Satkñnas, J., GrigienÅ, A., 2012. Eemian-Weichselian palaeoenvironmental re cord from the Mickñnai gla cial de pres - sion (East ern Lith u a nia). Geologija, 54: 35–51.
Satkunas, J., Grigiene, A., Jusiene, A., Damusyte, A., Mazeika, J., 2009. Mid dle Weichselian paleolacustrine ba sin in the Venta river val ley and vi cin ity (north west Lith u a nia), ex em pli fied by the Purviai out crop. Qua ter nary In ter na tional, 207: 14–25.
Satkunas, J., Grigiene, A., Buynevich, I.V., Taminskas, J., 2013.
A new Early-Mid dle Weichselian palaeoenvironmental re cord from a lac us trine se quence at Svirkanciai, Lith u a nia. Boreas, 42: 184–193.
Savvaitov, A., Straume, J., 1963. On the ques tion of twin strata in the mo raine of Valdai gla ci ation in the area be tween the lower reaches of the rivers Daugava and Gauja (in Rus sian with Eng - lish sum mary). In: Ques tions on Qua ter nary Ge ol ogy, 2 (ed. I.
Danilªns): 71–86. Acad emy of Sci ences of Lat vian SSR, RÌga.
Straume, J., 1979. Geo mor phol ogy (in Rus sian). In:
Geologicheskoye stroyenie i poleznye iskopayemye Latvii (eds.
J. Misªns, A. Brangulis, I. Danilªns and V. Kurës): 297–439.
Zinªtne, RÌga.
Svendsen, J.I., Alexanderson, H., Astakhov, V.I., Demidov, I, Dowdeswell, J.A, Funderf, S., Gataullin, V., Henriksena, M., Hjort, C., Houmark-Niel sen, M., Hubbertenk, H.W, Ingólfsson, Ó., Jakobsson, M., Kj³ri, K.H., Larsen, E., Lokrantz, H., Lunkka, J.P., Lys¯, A., Mangerud, J., Matiouchkov, A., Murray, A., Möller, P., Niessens, F., Nikolskaya, O., Polyak, L., Saarnisto, M., Siegert, C., Siegert, M.J., Spielhagen, R.F., Steins, R., 2004. Late Qua ter - nary ice sheet his tory of north ern Eur asia. Qua ter nary Sci ence Re views, 23: 1229–1271.
Thomas, D.S.G., Burrough, S.L., 2013. Lu mi nes cence-based dune chro nol o gies in south ern Af rica: analysis and in ter pre ta - tion of dune da ta base re cords across the sub con ti nent. Qua ter - nary In ter na tional, 21: 825–836.
Thrasher, I.M., Mauz, B., Chiverrell, R.C., Lang, A., 2009. Lu mi - nes cence dat ing of glaciofluvial de pos its: a re view. Earth-Sci - ence Re views, 97: 133–146.
Wallinga, J., 2002. On the de tec tion of OSL age over es ti ma tion us - ing sin gle-aliquot tech niques. Geochronometria, 21: 17–26.
Weckwerth, P., Przegiêtka, K.R., Chruœciñska, A., Pisarska-Jamro¿y, M., 2013. The re la tion be tween op ti cal bleach ing and sedimentological fea tures of flu vial de pos its in the Toruñ Ba sin (Po land). Geo log i cal Quar terly, 57 (1): 31–44.
Wintle, A.G., 2008. Lu mi nes cence dat ing: where it has been and where it is go ing. Boreas, 37: 471–482.
Wintle, A.G., Murray, A.S., 2006. A re view of quartz op ti cally stim u - lated lu mi nes cence char ac ter is tics and their rel e vance in sin - gle-aliquot re gen er a tion dat ing pro to cols. Ra di a tion Mea sure - ments, 41: 369–391.
Wyshnytzky, C.E., Rittenour, T.M., Summa, M., Thackray, G., 2015. Lu mi nes cence dat ing of late Pleis to cene prox i mal gla cial sed i ments in the Olym pic Moun tains, Wash ing ton. Qua ter nary In ter na tional, 362: 116–123.
Zelès, V., 1993. Glaciotectonic land forms of di ver gent type glaciodepressional low lands. Dis ser ta tion work syn the sis. Uni - ver sity of Lat via, RÌga.
Zelès, V., Markots, A., 2004. Deglaciation his tory of Lat via. In: Qua - ter nary Glaciations - Ex tent and Chro nol ogy. Part I: Eu rope (eds.
J. Ehlers and P.L. Gibbard): 225–243. Elsevier, Am ster dam.
Zelès, V., Markots, A., Nartišs, M., Saks, T., 2011. Chap ter 18:
Pleis to cene glaciations in Lat via. De vel op ments in Qua ter nary Sci ence, 15: 221–229.
490 Kristaps Lamsters, Edyta Kaliñska-Nartiša, Vitªlijs Zelès and Helena Alexanderson
