Stable isotope paleoaltimetry studies often implicitly assume that atmospheric flow interactions with topography can be simply modeled as a Rayleigh distillation process in which air parcels consistently ascend topographic barriers. We present a modern (A.D. 1979–2010) air parcel trajectory analysis using the Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model that shows that this fundamental assumption is often violated in the Sierra Nevada region of the western United States. Observed trajectory patterns and scaling calculations indicate that windward orographic blocking frequently occurs when trajectories encounter high elevations (>2.5 km) in the central and southern Sierra Nevada. As a result, trajectories reaching the Sierran lee commonly travel around, rather than over, the highest range elevations. Redirection effects are particularly pronounced at leeward sites distal (>150 km) to the Sierran crest, but are also evident in trajectory patterns in the northern Sierra Nevada. This trajectory analysis improves the interpretability of regional meteoric water and proxy isotopic records and has particular relevance to stable isotope–based reconstructions of Sierran paleoelevations. Specifically, stable isotope methods alone provide only limited insight into the elevation history of the Sierra Nevada and are likely insufficient to resolve proposed late Cenozoic elevation gains on the order of 1–2 km that may have raised the Sierra to its modern elevations.


The Cenozoic paleoelevation history of the Sierra Nevada (California, United States) remains ambiguous despite its importance for understanding the tectonic and geodynamic evolution of the region. In general, the debate centers on whether the Sierra Nevada is a long-lived topographic feature, or has experienced 1–2 km of late Cenozoic elevation gain (Fig. 1C) due to the well-documented removal of a dense lithospheric root from beneath the southern section of the range (e.g., Ducea and Saleeby, 1998; Jones et al., 2004). Geomorphic evidence, specifically rates and amounts of post-Eocene (Clark et al., 2005) and post–mid-Miocene (Unruh, 1991; Wakabayashi and Sawyer, 2001; Stock et al., 2004) fluvial incision, channel tilting, and/or base level change, is commonly cited in support of mid- to late Cenozoic elevation gain models. In contrast, regional paleo–meteoric water stable isotope (δD and δ18O) proxy records have been interpreted to suggest that the Sierra Nevada was a major topographic feature since at least the mid-Miocene (Poage and Chamberlain, 2002; Crowley et al., 2008) and may have been characterized by elevations comparable to modern as early as the Oligocene (Cassel et al., 2009) and Eocene (Mulch et al., 2006; Hren et al., 2010), thus calling into question how significant crust-mantle geodynamic processes (e.g., loss of eclogitic root) were for the topographic evolution of the Sierra Nevada.

At least part of the discrepancy among proposed paleoelevation histories may be attributed to along-strike variations in Sierran topographic development (e.g., Lechler and Niemi, 2011a), particularly in light of the fact that most Sierran paleo–meteoric water proxy records only inform about northern Sierra Nevada paleoelevations (north of ∼38°N; Mulch et al., 2006; Cassel et al., 2009; Hren et al., 2010). The exceptions are stable isotope paleoaltimetry studies using proxies collected from leeward sites that span nearly the full latitudinal range of the Sierra Nevada (Poage and Chamberlain, 2002; Crowley et al., 2008). These studies use observed trends in modern meteoric water isotopic distributions, where leeward δ18O values are ∼5–7‰ lower (δD ∼40–60‰ lower) than those at equal-elevation windward sites (Fig. 1A), as a basis for interpreting Sierran orographic barrier paleoelevations from leeward proxy isotopic records. Fundamental to this leeward proxy approach is the assumption that precipitating air parcels consistently travel up and over, and rain out heavy water isotopologues across, topographic barriers without significant blocking or redirection. This assumption neglects the dynamic interactions of atmospheric flows with topography (e.g., Smith, 1979), namely the potential for air parcels to be blocked by and redirected around high range-scale relief (>3 km) in the Sierra Nevada, which can complicate paleoelevation reconstructions from proxy isotopic records (Galewsky, 2009). The fact that modern precipitation δ18O (and δD) values are actually lower at the Sierran crest than in the downwind lee (Fig. 1A) suggests that such blocking and redirection effects are influential in the Sierra Nevada region and that air parcels precipitating at the Sierran crest are distinct from those reaching the Sierran lee.

Here we present a modern (A.D. 1979–2010) regional air parcel trajectory analysis to investigate how Sierran topography influences the pathways by which precipitating air parcels are transported through the region. Values for the nondimensional number Nh/U, where N is the Brunt-Väisälä frequency (a measure of atmospheric stability), h is the range-scale relief, and U is the cross-mountain wind speed far upstream of the Sierra Nevada (i.e., over the coastal Pacific Ocean), are also presented in order to discern among trajectories that are blocked (Nh/U > 1) from those that ascend Sierran topography relatively unimpeded (Nh/U < 1; Galewsky, 2009, and references therein).


The Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model (Draxler and Rolph, 2012) is a proven tool for investigating relationships between air parcel trajectories and resulting precipitation isotopic compositions (e.g., Sinclair et al., 2011; Bershaw et al., 2012). Here we use HYSPLIT to calculate air parcel back-trajectories (i.e., the pathway traveled) at locations with varying geographic positions relative to the Sierran orographic barrier (Fig. 1A). Trajectory analysis is presented for four sites in the lee of the high elevation (>2.5 km; Fig. 1) central and southern Sierra Nevada, both proximal (<30 km; Bishop and Lone Pine, California) and distal (>150 km; Tonopah and Beatty, Nevada) to the Sierran crest. Trajectory analysis is also presented for the mouth of the Yuba River and Reno, Nevada, which are located at the windward base and in the proximal lee of the northern Sierra Nevada, respectively (Fig. 1A).

We use North American Regional Reanalysis (NARR) climate data (http://www.esrl.noaa.gov/psd/) as the HYSPLIT climatological input. While NARR grid resolution (∼32 × 32 km) does smooth topography (Fig. 1B), primary regional topographic distributions are adequately resolved, making it sufficient to investigate topographic influences on air parcel trajectories at a regional scale. The NARR dataset also provides the relevant climate parameters required to calculate trajectory Nh/U values (temperature, specific and relative humidity for N; zonal and meridional wind speeds for U; see the GSA Data Repository1).

At each site (Fig. 1A), we calculated 5-day back-trajectories at 6 hr intervals for the 1979–2010 analysis period. In most cases, the 5-day duration was sufficient to trace trajectories to their oceanic source. Here we limit our analysis to trajectories that precipitated at the site of interest and only present results for a 1500 m trajectory start height, which is representative of trajectory patterns over the 1–2 km height range (see the Data Repository) that encompasses the atmospheric level where most water vapor is advected (e.g., Bershaw et al., 2012). The duration and dense sampling of our trajectory analysis reduces uncertainties arising from the fact that only air parcel trajectories, and not actual moisture transport pathways, are calculated by HYSPLIT (Helsen et al., 2007), and ensures a comprehensive representation of regional trajectory patterns.


The spatial patterns of air parcel trajectory pathways are presented in the form of trajectory contour plots (Fig. 2). Table 1 provides quantitative measures of how often trajectories crossed defined latitudinal domains in the Sierra Nevada region.

Contour plots (Fig. 2) show that actual air parcel trajectories in the Sierra Nevada region are more complex than the idealized “up-and-over” trajectories assumed by Rayleigh distillation models (e.g., Rowley and Garzione, 2007) and commonly applied in stable isotope paleoaltimetry studies (e.g., Galewsky, 2009). Observed regional trajectory patterns are clearly three-dimensional and directly influenced by Sierran topography (Fig. 2). Specifically, air parcels reaching leeward sites distal from the Sierran crest commonly are deflected around, and do not travel over, the high elevations (>2.5 km) in the central and southern Sierra Nevada (42% of air parcels arriving at Tonopah, Nevada, traversed the Sierra Nevada; 22% at Beatty, Nevada; Table 1), a finding consistent with published trajectory analyses for Winnemucca, Nevada, and Cedar City, Utah (Fig. 1A; Friedman et al., 2002). In contrast, the majority of trajectories reaching proximal leeward sites (Bishop and Lone Pine, California) do ultimately traverse the Sierran crest (79% at Bishop, 60% at Lone Pine; Table 1), even if trajectories are partially blocked or slowed upwind of the Sierra Nevada (Fig. 2).

Beatty trajectory patterns also highlight the importance of trajectory pathways that have little to no interaction with Sierran topography. Trajectories precipitating in Beatty regularly travel north from the Gulf of California (19%; Table 1) or enter the southern Great Basin through low-lying areas south of the Sierra Nevada (51%; Table 1) (Fig. 2). Similar southern trajectories also frequently deliver precipitation to Tonopah (46%) and Lone Pine (40%; Table 1).

Histograms of the maximum surface elevation crossed by each trajectory also highlight the disparity between proximal and distal leeward sites (insets, Fig. 2). Average maximum elevations traversed by trajectories arriving at Tonopah (2177 m) and Beatty (1639 m) are ∼500 m lower than those observed at similar latitude sites proximal to the Sierra (Bishop [2620 m] and Lone Pine [2216 m], respectively).

Precipitation arriving at the base of the windward northern Sierra Nevada (Yuba River) is generally derived from simple, west-east trajectories (Fig. 2). In contrast, south-to-north, range-parallel trajectories frequently deliver precipitation to higher-elevation (>1 km) sites within and in the lee of the northern Sierra Nevada (Reno; Fig. 2). This suggests that high (>2.5 km) elevations in the central and southern Sierra Nevada deflect and redirect trajectories northward through the Central Valley prior to trajectories resuming west-to-east transport directions upon reaching the moderate-elevation northern Sierra Nevada (∼2–2.5 km).


Trajectory Nh/U values provide a measure of the degree of blocking a trajectory will experience upon interacting with topography (e.g., Galewsky, 2009). Nh/U values <1 are associated with trajectories that are able to ascend topographic barriers, due to high cross-mountain wind speeds (U), low range relief (h), low atmospheric stability (i.e., low N), or some combination of these factors. In contrast, trajectories with Nh/U > 1 are commonly blocked by topography (Galewsky, 2009). Less than 20% of analyzed trajectories have calculated Nh/U values < 1 (h = 3 km; Table 1), which indicates that regional atmospheric stability (N) is too high and/or wind speeds (U) are too low for most modern trajectories to traverse the high relief (h) of the Sierra Nevada unimpeded. Consequently, as suggested by Galewsky (2009), Sierran stable isotope paleoaltimetry interpretations that assume low Nh/U flows (i.e., <1) are potentially flawed.


Modern (1979–2010) air parcel trajectories are strongly influenced by Sierra Nevada topography. Proximal leeward sites most frequently receive precipitation derived from air parcels that traverse the Sierran crest. As a result, meteoric water and proxy isotopic records collected from these locations have the greatest potential to be reliable recorders of Sierran elevation and paleoelevation distributions. In contrast, distal leeward sites receive a substantial portion of precipitation from trajectories that travel around high (>2.5 km) Sierran elevations (Fig. 2; Table 1). As a result, the assumption that Sierran lee precipitation is derived from air parcels that had previously rained-out over the presumed upwind Sierran crest is often invalid. This observation at least partially explains modern regional precipitation δ18O distributions (Fig. 1A) in which leeward δ18O values are higher than those at the Sierran crest. Consequently, modern and, presumably, paleo–meteoric water records collected at distal leeward locations are sensitive to only a portion of Sierra Nevada topography, limiting their utility for paleoelevation studies.

Latitudinal position with respect to the orographic barrier is also important. Beatty, Nevada, is located at the latitude of the highest Sierran peaks (∼36.5–37.5°N), but most trajectories reaching Beatty have little interaction with Sierran topography due to proximity to the low-elevation pathways accessible south of the range. This suggests that proxy records at leeward sites close to the northern and southern extents of the Sierra Nevada are especially sensitive to trajectories that travel around high topography and, as a result, provide limited information about windward elevations. The modern El Paso Basin, located east of the southernmost Sierra Nevada (Fig. 1A), is one such location that is highly susceptible to these wrap-around trajectories. Accordingly, interpretations of an ∼2 km decrease in southern Sierra surface elevations since the mid-Miocene, based on the observed ∼5–6‰ increase in δ18O values for ca. 14–6 Ma proxies collected from the El Paso Basin (Poage and Chamberlain, 2002), are unreliable. Instead, the observed increase in proxy δ18O values most likely reflects increased influence from trajectories traveling south of, rather than over, the Sierra Nevada.

Interestingly, northern Sierra Nevada proxy records potentially reflect some component of central and southern Sierra paleotopography, particularly once elevations in the southern part of the range exceeded ∼2.5 km. The frequency of blocking events associated with >2.5 km Sierran elevations also suggests that once the Sierra Nevada reached mean elevations of this magnitude, enhanced orographic blocking made leeward proxies relatively insensitive to subsequent increases in range elevation. As a result, leeward stable isotope paleoaltimetry methods will have difficulty resolving late-stage elevation gain on the order of 1–2 km, as has been commonly proposed to have resulted from loss of an eclogitic root during the Late Miocene–Pliocene (e.g., Jones et al., 2004).

Extrapolating the trends observed in this modern trajectory analysis to interpretations of proxy records must be done with caution, as changes in both climate (e.g., Eocene greenhouse; Molnar, 2010) and tectonic (e.g., opening of the Gulf of California) state likely influence trajectory patterns. Future work explicitly investigating how trajectory patterns change as a function of climate and tectonics will benefit interpretation of proxy records, but an understanding of modern trajectory variability when regional climate and topography are known, as presented here, is an essential calibration for any paleo-trajectory study. In addition, the common practice of using modern isotopic distributions to constrain past elevations (e.g., Poage and Chamberlain, 2002; Mulch et al., 2006) supports using modern trajectory analysis to clarify paleoelevation interpretations.


Sierra Nevada topography exerts a dominant influence on modern regional air parcel trajectories, with orographic blocking and redirection effects common in areas of high range elevations (>2.5 km). As a result, leeward paleo–meteoric-water proxy isotopic records frequently do not reflect the highest Sierran elevations (3–4+ km) and, instead, most reliably inform only about early stages of range growth when orographic blocking effects are relatively minor. The work presented here can at least partially reconcile contrasting Sierra Nevada paleoelevation histories. Low δ18O (and δD) proxy values in the lee of the Sierra Nevada are consistent with the presence of a moderate–high elevation (2–2.5 km) Sierran orographic barrier during the Middle Miocene (e.g., Poage and Chamberlain, 2002), but these proxies cannot rule out 1–2 km of elevation gain during the late Cenozoic (e.g., Jones et al., 2004) that would have brought the range to its modern elevations.

This research was supported by National Science Foundation (NSF) Experimental Program to Stimulate Competitive Research (EPSCoR) grant EPS-0918635, and NSF-Tectonics grant 1049903 (Galewsky). Two anonymous reviewers and Sandra Wyld are thanked for their reviews.

1GSA Data Repository item 2013063, contour plots for 1 and 2 km initialization heights, map of Sierran domains, and details about Nh/U calculations, is available online at www.geosociety.org/pubs/ft2013.htm, or on request from editing@geosociety.org or Documents Secretary, GSA, P.O. Box 9140, Boulder, CO 80301, USA.