151-18

MLDAWB

Richmond

PMD

MLD

PMDAWB

Lake Malawi

0

20

40

60

80

100

120

140

10

20

30

40

50

60

70

80 *

Paleosol

-1 0 1 2 3

*

*

**

* Shoreline ElevationConstraints

0 m -350m-100 -593m

5pt-Smoothed PC1wetter drierAGE

(ky) MBLF

0

5

10

15

20

25

30

35

40

MBLF

Density(gm/cm3)

1.0 2.2

0 6TOC %

Well-laminatedSilty Clay

Sparsely-laminatedsilty-clay

Grey, mottled,carbonate-richsilty clay

Sand

MAL 05-2AMAL 05-1CLake Malawi Synthetic

Stratigraphy

MLDAWB

MLD

AWB

MLD

AWB

Figure 9A: Wilkins Peak

PMD

SyntheticStratigraphy

MLDPMDAWB

Figure 10A: Laney 1

MLD

PMDAWB

Figure 10B: Laney 2

PMD

PMD

I propose a classification of lakes and their records based on three variables (Fig. 1, left):

MLD, maximum lake depth from floor to outlet.PMD, potential mean depth of the lake relative to MLD unconstrained by the outlet or lake floor; AWB, amplitude of water balance variability deviating from PMD

For illustrative purposes I have depicted AWB as a precession-related signal representing water input into the lake basin (with mpv being a time ofmaxiuum precessional variability), but it could have any form or not vary at all. If AWB = 0, this system yields the classification of Carroll &Bohacs (1999) lake types (1): “overfilled”, with PMD > MLD; “balanced fill”, with PMD = MLD; and “underfilled”, with PMD < MLD (Fig. 2, right).

However, AWB ≠ 0 in general, but fluctuates in frequency and amplitude, leading to a huge range of lake behavior and sediment sequence types,some examples of which are shown below. The first 3 examples are of early Mesozoic age and illustrate the classification with the three styles ofsequences seen in the rifts of Eastern North America (Olsen, 1990). Applications to a modern lake (Malawi) and the Eocene Green River Fm. showthe versatility of the method. This method has some similarities to that of Keighley et al. (2003).

New Dynamic Classification of Lake Systems and Their Geological RecordsPaul E. Olsen (LDEO, Palisades, NY: polsen@LDEO.columbia.edu)

MLD

AWB

mpv

PMD

height of outlet

lake floor

MLD

PPLLDDAWB = 0

Balanced Fill lake type of Carroll & Bohacs (1999)

PMD = MLD = outlet

MLD

AWB = 0 PPLLDD

PMD > MLD

Overfilled lake type of Carroll & Bohacs (1999)

MLDPPLLDDAWB = 0

PMD < MLD

Underfilled lake type of Carroll & Bohacs (1999)

Figure 1: Basic definitions (see right).Figure 2: Lake types of Carroll & Bohacs (1999) in terms of new classification (see left).

Md An Cms Bs Sy Ca

545.8

1m

Basalt

Md, mud; An, anhydrite; Cms, chaotic mudstone-salt;Bs, banded salt; Sy, Sylvite; Ca, Carnellite

Core AV-C-1-4Fundy basin, Nova Scotia

Core Pz-18Kehmisset basin, Morocco

AV-C-1-4 (730-780 ft) spanning basal North Moun-tain Basalt (base at arrow) and the uppermostBlomidon Fm. Core is youngest in upper right.

SyntheticStratigraphy

Figure 3: Fundy-type lacustrine sequences: A, (below), relationship between vari-ables; B, (left-top) salt dissolution autobreccia bowls, Fundy basin, Nova Scotia, Can.;C, (left-bottom) sand patch cycles, also from the Fundy basin.

Figure 3AFigure 3B

Figure 3C

49.62±0.10

49.79±0.09

49.96±0.08

50.39±0.13

50.61±0.26

50.83±0.13

51.30±0.30

49.5

50.0

50.5

51.0

512

256

128

64 32 16 8 4 0 10 20 30 40

40Ar/39Ar AgesSmith et al.,

2008

Cycle AgeMachlus et al.,

2008PERIOD (ky) OIL YIELD (gal/ton)

El Paso 44-3

400 100 20 kySynthetic

Stratigraphy

SyntheticStratigraphy

In the case of Fundy, type sequences PMD was near or below the base of MLD, and deepening excursions of the AWB is less than the top of MLD. As a consequence, thelake and its record may be dominated by short periods of high-stands at mpv, and fre-quent periods of evaporite production, and long periods of eolian deposition on salineflats (sand-patch cycles: Fig. 3C). In many outcrops and at the basin margins the evap-orite layers may be represented by hiatuses or solution autobreccias (Fig. 3B). Exam-ples of the lateral transitions from mud-dominated cycles at the fringes of basins toevaporite-dominated sequences are common in the eastern North America and Moroc-co. These closed saline basins, because of their nearly continuously negative waterbalance can receive seepage from brines of marine origin if they are available.

Figure 3D

Figure 4: Cores of Fundy-type sequences below the oldest basalt in theFundy and Kehmisset basins. Marginal facies of the salt-rich deep parts of thelatter resemble the former.Newark

Fundy

Figure 5: Newark-type sequences: A, (above-right),relationship between variables; B,~ 6 m thick cycles(black to black) tracking climatic precession in theLate Triassic, lower Lockatong Fm, Newark basin,PA; C, ~11 m cycles, (black to black layers, track-ing precession), Early Jurassic East Berlin Fm.,Hartford basin, CT. Black layers are organic-richand microlaminated with whole fish.

Figure 5A

Figure 5B Figure 5C

In Newark-type sequences PMD was slightly above the base of MLD, and deep excursions of the AWB higher than the top of MLD during mpv (Fig. 5A). Deepestlakes occur at mpv flushing solutes out, but then drying up. Extreme dry times do notleave a distinct record compared to times of low precessional variability, and can becharacterized by non-deposition or deflation. High eccentricity (mpv) leave a recordof bundles of high-TOC cycles separated by low-TOC cycles. Dispersed evaporitested to occur in high amplitude cycles that do not reach the top of MLD (Fig. 5D).

Figure 7B

Figure 6: Cores of Newark-type sequences from the Newark Basin. Left,Lockatong Fm. (1620-1644 ft, Nursury no. 1); lake-level (precession) cycle isfrom black to black layer. Right, Passaic Fm. (808-835 ft, Somerset no. 1);lake-level cycle is banded silts (~815.5 ft) to banded silts (~823 ft) .

PMD was well above the base of MLD, in Rich-mond-type sequences, with deepening excur-sions of the AWB higher than the top of MLDand lower than MLD at times of mpv (Fig. 7A,left). In this mode, both deepest lake and shal-lowest lake sequences occur at mpv, with thelake rarely drying out. If the deepest-water unitsare similar in facies to the mean facies, it is easyto get the phase of the eccentricity cycle wrongrelative to Newark-type sequences.

Figure 5D

Figure 7: (Right) Richmond-type sequences: A,(above-right), relationship between variables; B,exposure of lower Vinita Fm., Richmond basinwith one cycle of microlaminated black shaleand gray sandstone; C, Vago no. 1 core showingalternating better laminted and more poorly lam-inated intervals, all with high TOC (>2%).

Figure 7A

Figure 7C

Figure 8: Lake Malawi and its sedimentary record: A, model showing relationship ofthe classification variables; B, synthetic stratigraphy derived from A; C, principle com-ponent 1 (PC1) of various biotic and physical properties of core MAL 05-1C that areproxies of climate and hence water depth (modified from Cohen et al., 2007) correlatedto the synthetic stratigraphy (B); D, core MAL 05-2A from a more up-dip position, withan unconformity corresponding to one of the shallower excursions in C and showingexample of sedimentary features of the core (modified from Scholz, et al., 2007).

(Left) In terms of this new classification, Lake Malawi is characterized by having PMD above the top of MLD and an AWB that is larger than MDL. The dry excursions of AWL, infrequently pass below the base of MLD (Fig. 8A). Thus despite a large amplitude climate forcing, because PMD is above the top of MLD, the lake is usually hydrologically open. However, during times of high mpv, there are large lake level drops, resulting in erosion shoreward and carbon-ate deposition, and rare exposure and paleosols develop-ment, basinward (Fig. 8C). As a consequence the sedimen-tary record consists mostly of deeper-water deposits, punctuated by shallow water excursions (Fig. 8B-8D). These shallow-water excursions are the megadroughts of Scholz et al. (2007) and Cohen et al. (2007).

In many ways, the Malawi sequence resembles the Rich-mond-type sequences, except that Malawi is even more per-sistently deep as indicated by PMD being above the top of MLD.

A counter intuitive (at first) consequence of this relation-ship is that the intervals of the core with the most indica-tions of aridity and major drops in lake level, are possiblealso those that experienced some the the most extremehumid events, uncoupled however from increases in lakedepth.

The Green River Fm. arguably constitutes the most famous lacustrine deposit inthe world. The Wilkins Peak Mb. of the formation in Wyoming has in addition tooil shale, cyclicaly disposed (Fig. 9C), significant beds of evaporites (e.g., trona)and many clear surfaces of exposure. The overlying Laney Mb, has a much higherconcentration of oil shale and few or no surfaces of exposure. The sequence isprofoundly cyclical and astronomically tuned records of oil shale yield (Machluset al., 2008) produce a chronology remarkably close to 40Ar/39Ar chronology ofSmith et al. (2008) (Fig. 9C).

In this classification the Wilkins Peak has PMD very close to the base of MLD with wet excursions of AWB never reaching MLD (Fig. 9A), and hence, there isno solute flushing.

The new classification suggest two ways the transition from Wilkins Peak to Laney may have occurred. One is by effectively lowering the outlet (Fig. 10A). The other is by raising PMD above the top of MLD (outlet) by either climate change or watershed capture (Fig. 10B). The two mechanisms suggest rather dif-ferent synthetic stratigraphies. Which is a better fit to reality?

Blacks Fork no. 11 m

538-553

Figure 9: Wilkins Peak Mb. of the Green River Fm.: A, rela-tionship of the variables; B, Wilkins Peak overlain by LaneyMb.; C. Synthetic stratigraphy; D, Wavelet Spectrum of theWilkins Peak from Machlus et al. (2008: cycle age at 0 onMachlus, reassigned a value corresponding to the uppermostdate of Smith) with the ages of Smith et al. (2008) superim-posed; E, portion of Blacks Fork no. 1 core, showing alterna-tions of high (1) and low (2) TOC interals.

SyntheticStratigraphy

SyntheticStratigraphy

References: Carroll AR & Bohacs KM. 1999. Geology 27:99; Cohen AS et al. 2007. PNAS 104:16422-16427; Keighley D. et al. 2003. Jour. Sed. Res. 73(6):987-1006; Olsen PE. 1990, in Katz B (ed), LacustrineBasin Exploration, AAPG Memoir 50:209-224; Machlus ML. et al. 2008. EPSL 268:64-75; Scholz CA et al.2007. PNAS 104(42):16416-16421; Smith ME et al. 2008 Nature Geoscience 1(6):370-374.

My Thanks to Malka Machlus for Fig. 9E and Mohammed Et-Touhami for Fig. 4(right).

Figure 9B

Figure 9CFigure 9D Figure 9E

Red mudstone

Gray mudstone

Dark Graymudstone

Black mudstone

Range of mudstonecolors with evaporitecrystals

Bedded evaporites

Key to SyntheticStratigraphies