For orbital tuning, we considered shipboard core logging data (magnetic susceptibility, bulk density, and color spectra) as well as carbonate percentages, sand fraction percentages of the carbonate fraction, and benthic isotope records. Core logging shipboard data (Mix, Tiedemann, Blum, et al., 2003) were measured at intervals of 2.5 cm (Site 1237) and 5 cm (Site 1241) and provide a temporal resolution of 1000–2500 yr, which corresponds to sedimentation rates ranging from 1 to 3 cm/k.y. at Site 1237 and from 2 to 5 cm/k.y. at Site 1241. Isotope data, carbonate, and sand percentages were measured every 5 cm at Site 1237 and every 10 cm at Site 1241.
Isotope data from Sites 1237 and 1241 were analyzed at the IFM-GEOMAR Institute in Kiel (Germany) using a Finnigan/MAT-252 mass spectrometer equipped with a fully automated Finnigan/Kiel-Carbo-II carbonate preparation device. For each isotope analysis, up to five specimens of the epibenthic foraminifers Cibicidoides wuellerstorfi or Cibicidoides mundulus were picked from the 250- to 500-µm
fraction. The laboratory standard was calibrated to Peedee belemnite through National Bureau of Standards reference material NBS-19. Analytical reproducibility of the laboratory standard typically was about
±0.06‰ for 
18O and ±0.03‰ for 13C (±1
). Isotope data from Site 1236 were measured at the Leibniz Labor for Radiometric Dating and Isotope Research at Kiel University (Kiel, Germany) using a Finnigan/Delta-Plus-XL mass spectrometer coupled to a Finnigan/Gas-Bench-II. Precision of the local carbonate standard was
±0.07‰ for 18O and ±0.05‰ for 13C (±1) over the period of analyses.
The CaCO3 contents were determined by infrared absorption of total CO2 (organic and inorganic carbon) released by combustion with a LECO analyzer. Early Pliocene organic carbon contents are lower than 0.4 wt% at Sites 1236, 1237, and 1241 (Mix, Tiedemann, Blum, et al., 2003). Hence, carbonate percentages might be overestimated by as much as 3.3 wt% (equal to an organic carbon content of 0.4 wt%).
The sand fraction represents the wet-sieved residue of the >63-µm fraction, which consists of nearly 100% carbonate. The content of the dry sand fraction is given as weight percent of total CaCO3 to compensate for dilution effects caused by noncarbonate dilutants.
For orbital tuning and spectral analyses, we use the software package AnalySeries 2.0 (Paillard et al., 1996). AnalySeries especially facilitates the transformation of "proxy data vs. depth" records into "proxy data vs. age" records by graphically adjusting cyclic fluctuations of paleoclimatic proxy records to the astronomical target record. The age of each data point was estimated by linear interpolation between age-depth control points. AnalySeries also provides filters and a set of spectral analysis methods that are partly complementary in terms of robustness vs. resolution. We preferred to use the classical Blackman-Tukey method (Blackman and Tukey, 1958) for spectral analysis in the time and depth domain. The algorithm computes first the autocovariance of the data, applies a window, and finally applies a Fourier transform to compute the spectrum. It is a very robust method, unlikely to present spurious spectral features. The main drawback is its poor resolution in the spectral domain because sharp features can be considerably smoothed. This method allows the user to choose a resolution vs. confidence parameter: the length of the autocovariance series, which is generally set to 60% of the input series. The confidence level associated with the error bar on the spectrum is typically set to 80%. For filtering and spectral analysis in the time domain, we interpolated each record at equidistant intervals, corresponding to the average time resolution of the proxy record. Tukey bandpass filters with a central frequency of 0.045 cycles/k.y. (bandwidth = 0.01) and 0.024 cycles/k.y. (bandwidth = 0.009) were used to extract the precession- and obliquity-related components from the proxy records, respectively.
We used the orbital solution from Laskar et al. (2004) for Pliocene variations in Northern Hemisphere summer insolation (June 21–July 21), obliquity, and precession as a tuning target. Until 1996, orbitally derived age models were based on the astronomical solution of Berger and Loutre (1991). Lourens et al. (1996) demonstrated, however, that unrealistic large time lags will occur between obliquity and the obliquity-related variations in the proxy records if the orbital data from Berger and Loutre (1991) are used as a tuning target. Lourens et al. (1996) pointed out that the geological record can be calibrated most accurately to the summer insolation record calculated from the Laskar et al. (1993) solution La93(1,1) with a dynamically ellipticity of the Earth of 1 and a tidal dissipation term of 1, both close to present-day values. The new solution La2004 for the astronomical computation of the insolation quantities on Earth (Laskar et al., 2004) improved the solution La93 by using a direct integration of the gravitational equations for the orbital motion and by improving the dissipative contributions, in particular in the evolution of the Earth–Moon System. This orbital solution has been used for the latest calibration of the astronomically tuned Neogene timescale (ATNTS2004; Lourens et al., 2004) and is expected to be used for age calibrations of paleoclimatic data over the last 40–50 m.y.
A major precondition for tuning marine paleoclimate records to the astronomical record is to ensure the recovery of complete and undisturbed sediment records. This is achieved by drilling multiple offset holes at the same site location to splice across coring gaps and distorted sediment sequences through interhole correlations using closely spaced core logging measurements. This strategy (Ruddiman et al., 1987) became a standard during paleoceanographic ODP legs. At Sites 1236, 1237, 1239, and 1241, the construction of the composite depth was based on core logging magnetic susceptibility and GRA density data. We reinspected the composite depth of Sites 1237 and 1241 by considering in addition the core logging color data, which was not possible during the cruise because of time constraints. We did not verify the composite depth at Sites 1236 and 1239.
Only at Site 1241, the correlation of overlapping sections from adjacent holes by means of color data suggests a mismatch between Cores 202-1241C-11H and 202-1241A-16H (Fig. F1). The former correlation resulted in a doubling of a 63-cm-long sediment section. Our new splice suggests a switch point at 175.77 meters composite depth (mcd), from Section 202-1241C-11H-7, 20 cm (same as before), to Section 202-1241A-16H-4, 72 cm. This deletes 63 cm from the former composite depth (interval 202-1241A-16H-4, 7–72 cm; corresponding to an interval of 65 cm on the meters below seafloor scale). The revision leads to a small reduction of the composite depth scale, as 63 cm is subtracted from the mcd below Section 202-1241A-16H-4, 72 cm.
At best, the tuning medium should be marked by cyclic, high-amplitude fluctuations and a high signal-to-noise ratio. From our experience, however, we know that a single climate proxy record often includes an interval of low signal-to-noise ratios, where a clear interpretation of the geological record with regard to its orbitally induced variability is difficult to derive. Therefore, we used a multiproxy approach for orbital tuning to overcome such problems. Discrete ash layers (Mix, Tiedemann, Blum, et al., 2003) were removed from the data sets to avoid misinterpretations of the climate signal and distortions of sedimentation rates. The sediment record from Site 1237 was especially affected by the frequent deposition of ash layers. Fourteen ash layers were identified in the time interval from 2.4 to 6 Ma by the Leg 202 shipboard party, some of them ranging in thickness from 15 to 36 cm. Inspection of the sand fraction record indicated that the range with significant amounts of ash often spanned a larger interval than simply suggested by a discrete ash layer. Although we deleted discrete ash layers from the composite depth, the "normal pelagic sedimentation rate" is overestimated by the amount of dispersed ash in such intervals.
The first step toward astronomical calibration was to identify dominant cycles of selected proxy records in the depth domain by means of spectral analyses on succeeding 10-m-long intervals. This approach helped to trace long-term changes in sedimentation rates (e.g., an increase in sedimentation rate would move dominant precession- or obliquity-related frequencies [cycles/m] to lower frequencies). The spectral comparison between different proxy records also clearly identifies precession- and obliquity-related cycles and indicates which of the proxy records are best suited for tuning to the precession and/or obliquity.
In a second step, we extracted the precession- and obliquity-related components from the proxy records by bandpass filtering each of the succeeding 10-m-long sections. We then used the merged filter outputs for a first tuning. We started with tuning to obliquity for two reasons. First, most proxy records revealed a stronger response to obliquity than to precession. More than one proxy for tuning to obliquity improved the continuous and reliable detection of the 41-k.y. cycle from 2.5 to 6 Ma. Second, the effect of obliquity on insolation is symmetric across hemispheres. That is, cold summers occur in both hemispheres at the same time (in phase). This increases the likelihood that proxy records indicative of global climate change respond to obliquity forcing with possibly different but relatively constant phase lags. Pliocene variations in benthic 18O and 13C are dominated by 41-k.y. cycles and are thought to respond with relatively constant phase lags to obliquity forcing. Fluctuations in benthic 18O records are linked to high-latitude processes because they are sensitive to variations in global ice volume and to changes in deepwater temperature/salinity. Provided that the effect of obliquity on insolation is symmetric across hemispheres, the mid-Pliocene shift from unipolar to bipolar glaciations or changes in the location of predominant deepwater formation (Southern Ocean vs. North Atlantic) should not have significantly affected the phase of the obliquity-related benthic 18O variations (Clemens, 1999). The work of Imbrie and Imbrie (1980) and Imbrie et al. (1984) suggested a lag between obliquity and obliquity-controlled variations in ice volume of 8 k.y. for the Pleistocene time interval. Their model also implied that the time lag strongly depends on the size of the ice sheet, whereas a smaller ice sheet would result in a smaller phase lag. For the relatively warm Pliocene interval prior to Northern Hemisphere glaciation, the true phase lag may have been close to 6 k.y., as suggested by Chen et al. (1995). The disadvantage of solely using Pliocene benthic 18O records for orbital tuning is their low signal-to-noise ratio because the small variability in global ice volume prior to Northern Hemisphere glaciation (>3 Ma) distinctly decreased the 18O amplitudes. In some intervals of the early Pliocene, the signal-to-noise ratio is so low that it is difficult to decide whether 41-k.y. cycles are registered or not (Tiedemann et al., 1994; Shackleton et al., 1995). In this context, the benthic 13C records are an alternative for tuning to obliquity. Obliquity-related fluctuations in benthic 13C are thought to be largely controlled by global variations in marine productivity and the mass of organic matter stored in forests, soils, and shallow marine sediments (Shackleton, 1977), presumably related to glacial–interglacial climate change. Although the globally correlative nature of the Pliocene benthic 13C signal has never been examined in detail, several studies indicated that Pliocene benthic 13C maxima lag 18O minima in the eastern Pacific with a relatively constant phase of ~2 k.y. at the obliquity band (Mix et al., 1995; Shackleton et al., 1995). Considering a phase difference of 6 k.y. between Pliocene variations in orbital obliquity and benthic 18O, 13C maxima may have lagged obliquity maxima by ~8 k.y. For an unambiguous tracing of 41-k.y. cycles, we also considered other proxy records like GRA density, magnetic susceptibility, and color reflectance data. After determining their phase relationships with respect to 18O minima (warm stages), we established a preliminary age model that is based on tuning their 41-k.y. components to orbital obliquity. At this stage, the tuning provided constant phase relationships between the proxy records and orbital obliquity but did not include the possible phase lags for 18O and 13C, as mentioned above. Instead of that, we continued with tuning to precession; matching the obliquity-related filter output of a proxy record to orbital obliquity is not as easy as tuning to precession because the amplitudinal variability in orbital obliquity is low and thus provides no eye-catching control points for a correct match.
The advantage of tuning to orbital precession is that the amplitudinal power of orbital precession is well structured by eccentricity-induced packets of 100- or 400-k.y. intervals. These packets are often easily recognized in proxy records that are dominated by precessional variance (e.g., Tiedemann et al., 1994; Tiedemann and Franz, 1997). Therefore, we also used such packets for aligning climate variables with strong precessional responses. GRA density data from Site 1237 and the sand fraction record from Site 1241 reveal significant response to orbital precession forcing. Their precession-related filter outputs provided a powerful independent check on the correlation initially defined by tuning to obliquity. Nevertheless, we are aware of the fact that the amplitudinal variability of precession-related cycles could also be influenced by changes in sedimentation rates or internal processes of the climate system. Low sedimentation rates decrease the time resolution and the signal-to-noise ratio and thus may distort the original, amplitudinal climate response to orbital forcing. Internal processes of the climate system couple specific variables or mutually interact among them. These interactions may either amplify anomalies of one of the interacting elements or damp them. Thus, we only set precession-based age control points, where the alignment with the orbital record is corroborated by both the precession component and the obliquity component of proxy records. A point of discussion is that we have no model linking the precessional response of the proxy records to orbital forcing. Precession insolation forcing is hemispherically asymmetric with Northern Hemisphere summers being 180° (11.5 k.y.) out of phase with Southern Hemisphere summers and we did not examine whether Northern or Southern Hemisphere insolation exerted a stronger control on the precessional signal of the proxy records. We simply assumed that the observed correlation between benthic 18O warm stages and obliquity-related maxima or minima in proxy records is similar for precessional minima. For instance, if obliquity-related maxima in GRA density correlate with 18O warm stages (maxima in orbital obliquity), we also assumed that precession-related maxima in GRA density correspond to precession-controlled maxima in insolation. We used the summer insolation at 65°N as a target record and assumed an in-phase relationship between insolation and the proxy record. However, we did not tune to precession if the adjustment would lead to unrealistic large time lags between obliquity and glacial-bound variations in proxy records indicative of global climate change.
Sedimentation rates are relatively low (<3 cm/k.y.) at Site 1237. For this reason, we first demonstrate that the quality of the logging data from Site 1237 is still appropriate for orbital tuning (Fig. F2). The depth interval from 60 to 80 mcd is especially suited for a test because it has a well-constrained paleomagnetic age model that provides a first approximation of changes in sedimentation rates. Sedimentation rates range from 1.6 to 2 cm/k.y. between the base of Mammoth Chron and the base of Cochiti Chron. Within this interval, cyclic fluctuations of the GRA bulk density and magnetic susceptibility records are characterized by two major clusters of cycles, from 37 to 49 cm and from 77 to 114 cm (Fig. F2). These two clusters clearly identify the response to precession (19- and 23-k.y. cycles) and obliquity (41-k.y. cycle), whereas the range of the clusters is defined by the variability of sedimentation rates.
Spectral analyses in the depth domain suggested that the records of benthic 13C, magnetic susceptibility, and GRA density are best suited for orbital tuning. The 13C and magnetic susceptibility records revealed significant variability at the obliquity frequency band. The GRA density record provided in addition significant variability at precession-related frequencies (Fig. F2). Before tuning, we examined their relationships to obliquity-dominated changes in global climate as indicated by comparisons with the benthic 18O record. The GRA densities vary according to the nature of the sedimentary matrix, which is dominated by carbonate and smaller amounts of siliciclastics and biogenic opal. Maxima in GRA density reflect maxima in carbonate percentages. Obliquity-related maxima in GRA densities are associated with 18O warm stages and 13C maxima. The magnetic susceptibility record is negatively correlated to the carbonate record and represents the degree of magnetized sediment and, hence, approximates the ratio of siliciclastic vs. biogenic material. Maxima in magnetic susceptibility reflect relatively higher amounts of siliciclastics during 18O cold stages. The siliciclastic fraction is mainly derived from eolian deposition, as Site 1237 underlies the modern path of dust that originates in the Atacama Desert of Chile (Mix, Tiedemann, Blum, et al., 2003). Accordingly, we correlated obliquity-related minima in magnetic susceptibility, maxima in GRA density, and maxima in benthic 13C with maxima in orbital obliquity (warm stages). We involve the benthic 18O record into this process in those intervals where obliquity-related cycles were clearly registered. We started our tuning from the Gauss/Matuyama boundary (47.9
± 0.1 mcd), which has a well-constrained astronomical age of 2.59 ± 0.01 Ma (Shackleton
et al., 1995; Lourens et al., 1996). Other magnetic reversal boundaries were not considered during the tuning process. An initial age model for the interval from 2.1 to 6 Ma was derived by tuning the 41-k.y. component of the magnetic susceptibility record to orbital obliquity. The tuning was then verified via comparisons with the obliquity-related GRA density filter output. In a final approach, we matched the precessional component of the GRA record to the insolation record (Fig. F3) and reduced the number of tie points to an amount sufficient to keep the GRA density record approximately in phase with obliquity and precession (Table T1).
After tuning, we extracted the precession- and obliquity-related components from the proxy records by using Tukey bandpass filters. The filtered components are compared with the respective orbital time series in Figures F3 and F4. The 21-k.y. component of the GRA density record reveals a remarkably good resemblance with orbital precession in the intervals from 2.2 to 4.7 Ma and from 5.6 to 6.0 Ma, in particular with respect to the eccentricity modulation as reflected in the amplitude variations. In addition, a very consistent relation is found between the 41-k.y. components in our proxy records (GRA density and benthic 13C and 18O) and astronomical obliquity. The age model from 4.7 to 5.6 Ma is mainly based on correlating the 41-k.y. signal of the isotope records to obliquity (Fig. F4) because the GRA density record provided a weak variability at the obliquity band, especially between 4.9 and 5.5 Ma, and no clear similarity in amplitude fluctuations between orbital precession and the 21-k.y. GRA component.
In addition, a convincing age model should not produce physically unreasonable changes in sedimentation rates, especially in pelagic regions. The sedimentation rates at pelagic Site 1237 were found to vary from 1.1 to 2.8 cm/k.y., which seems reasonable. This range provides no evidence for distortions caused by the age model (Fig. F3) and is close to the values obtained from initial bio- and magnetostratigraphy.
To further examine our timescale, we used cross-spectral analyses to determine the time lags and coherencies of the proxy records with respect to the orbital time series. Coherency between the geological data and the orbital target in the precession band is one of the fundamental methods by which a timescale may be evaluated (Shackleton et al., 1995). Separating the time interval from 2 to 6 Ma into four succeeding intervals, cross-spectral comparison indicated that Northern Hemisphere summer insolation and GRA density were in phase (as requested by our tuning) and displayed high coherencies at the precession (mainly >0.95) and obliquity (0.91–0.97) frequency bands (Table T2; Fig. F5). The benthic 13C record provided coherencies of 0.97 and 0.89 for the time intervals from 4.2 to 5 Ma and from 5 to 5.9 Ma, respectively. Brüggemann (1992) showed that high coherencies are very unlikely to appear by tuning a randomly fluctuating time series to the astronomical record. Thus, the very high coherencies at Site 1237 would not be obtained without a close coupling between changes in insolation and southeast Pacific paleoceanography. Cross-spectral analyses also indicated fairly constant phase relationships between variations in orbital obliquity and benthic isotopes. For consistency of phase calculations, the 18O record was multiplied by –1, so that larger values indicate interglacial conditions along with maxima in Northern Hemisphere summer insolation. Variations in benthic 13C lag obliquity forcing by ~5 k.y. and obliquity-related variations in –18O by ~2 k.y. Accordingly, the benthic –18O signal, indicative of changes in ice volume, would lag obliquity forcing by ~3 k.y. As a consequence, the tuned ages at Site 1237 might be a few thousand years too old rather than too young, if the true phase lag was close to 6 k.y., as suggested by Chen et al. (1995).
The proxy records from Site 1241 contain significant distribution of variance at obliquity-related wavelengths of 100–200 cm. Accordingly, sedimentations rates vary from 2.5 to 5 cm/k.y. and are higher than those at Site 1237. The primary signals that we used for tuning were benthic 13C and sand percentages of the carbonate fraction. Both records are highly coherent with 18O at the obliquity frequency band. Maxima in 13C and sand percentages are associated with 18O warm stages. The sand fraction record is additionally marked by strong concentration of variance at precessional frequency bands. Carbonates and GRA density values show distinctly weaker correlations to 18O, although obliquity-related minima in GRA density and carbonate percentages are mainly associated with 18O warm stages. This correlation is opposite to that found at Site 1237. These data were only considered for identifying obliquity- or precession-related cycles in intervals where the variance or the response to orbital forcing was less pronounced in the 13C and sand fraction records.
Before using the sand fraction data for tuning, we examined the type of paleoceanographic information provided by this record. Variations in the sand percentages of the carbonate fraction are often used as a proxy to reconstruct either changes in carbonate dissolution or differences in carbonate accumulation between foraminifers and calcareous nannoplankton. The sand fraction consists of nearly 100% of foraminiferal shells at Site 1241. For such environments, it has been shown that the sand content of deep-sea carbonates decreases as dissolution progresses (e.g., Bickert et al., 1997). Foraminiferal shells are weakened by dissolution and tend to break down in small fragments. As a result, material moves from the coarse fraction into finer fractions. At the shallow water depth of Site 1241 (2027 m), carbonate dissolution is expected to have been low. Therefore, it is not surprising that the Pliocene variability in sand fraction percentages does not correspond to the general pattern of Pacific carbonate dissolution. Over the past 4.5 m.y., carbonate dissolution in the equatorial Pacific was enhanced during interglacials (Farrell and Prell, 1991). Thus, Pliocene changes in carbonate dissolution mainly operated on 41-k.y. cycles. These relationships rather exclude changes in carbonate dissolution as a primary mechanism because high sand percentages at Site 1241 are associated with interglacials and reveal in addition strong precession-related variability. As carbonate dissolution seems to be of secondary importance, changes in the ratio between nannofossil placoliths and foraminiferal shells are regarded as the primary mechanism that could change the relative portion of the coarse fraction (Bickert and Wefer, 1996). Hence, the sand fraction record is indicative of changes in carbonate productivity.
We used the oxygen isotope Stages 96, 98, and 100 as a starting point for calibrating the timescale at Site 1241. These stages were easily recognized in the benthic oxygen isotope record (Fig. F6). Their age assignments to the orbital record are well constrained (e.g., Tiedemann and Franz, 1997) and identified the three corresponding obliquity cycles. We then tuned the 41-k.y. component of the benthic 13C record to orbital obliquity assuming no phase difference. This resulted in a preliminary timescale for the interval from 2.4 to 5.7 Ma. In a next step, we compared the sand fraction record with the Northern Hemisphere summer insolation record and tuned the precession-related maxima of the sand record to insolation maxima (Fig. F6). The new age model changed the phase relationship between the 13C record and orbital obliquity with 13C lagging obliquity insolation forcing relatively constant by ~5 k.y. and lagging 18O by ~3 k.y. during the interval from 2.5 to 5.5 Ma (Table T2). Mix et al. (1995) found a similar phase relationship for the Pliocene between benthic 13C and 18O at tropical east Pacific Site 849 (2 k.y.). Results from cross-spectral analyses from Site 1241 imply that the benthic 18O signal, indicative of changes in ice volume, lagged obliquity forcing by ~2 k.y. For comparison, the tuning at Site 1237 yielded a similar phase lag of ~3 k.y. The tuned ages at Site 1241 might be a few thousand years too old rather than too young if a larger phase lag close to 6 k.y. is assumed (Chen et al., 1995). The age-depth control points for Site 1241 are given in Table T3.
To evaluate our tuned age model and the match between the proxy records and the orbital record, we applied the same methods used for Site 1237. Tuning to precession resulted in a concentration of variance over all the main orbital frequencies. For statistical evaluation of the tuned timescale, we applied cross-spectral analyses to estimate the coherencies between the proxy records and the orbital target for the time intervals from 2.5 to 3.5 Ma, 3.5 to 4.5 Ma, and 4.5 to 5.5 Ma (Fig. F7). At the obliquity band, coherency estimates for the sand fraction and benthic 13C and 18O records are >0.94. At the precession bands, the sand fraction record provided coherency estimates of >0.9. Table T2 summarizes the coherencies and phase estimates of the different proxies for Site 1241. The high coherencies also indicate that the physical linkage between changes in solar insolation and paleoceanography was strong through the entire interval from 2.5 to 5.5 Ma. Sedimentation rates were estimated and vary between 2 and 5 cm/k.y. Higher sedimentation rates mark the older part of the record. This is reasonable because the depositional environment in the equatorial east Pacific was characterized by a late Miocene to early Pliocene biogenic bloom, enhancing the sediment flux to the seafloor. Finally, we applied bandpass filtering to extract the orbital frequency components from the sand fraction, 13C, and 18O time series and compared them with orbital precession and obliquity. We found a remarkable similarity between the amplitude variation in orbital precession and the precessional components in the sand fraction record (Fig. F6). This implies that we correctly mapped the climate signal onto the orbital record. A mismatch of precession-related cycles (e.g., making one precession cycle too old or too young) would be recognized by an out-of-phase relationship with orbital obliquity. The filtered 41-k.y. components in the sand fraction, 13C, and 18O records show a very consistent relationship with orbital obliquity, with different but relatively constant phase offsets (Table T2). The mismatch between orbital obliquity and the 41-k.y. component of the sand fraction at ~2.9 Ma is caused by relatively high amounts of ash, although we deleted the corresponding interval of the ash layer (77.08–77.10 mcd) (Mix, Tiedemann, Blum, et al., 2003) before tuning. Inspection of the sand fraction, however, revealed relatively high amounts of ash over an interval of ~1.5 m (76.27–77.77 mcd) that clearly distorted the primary signal of the sand fraction record.
