2005 Thesis Abstracts

Harumi Goya, 2005, Stratigraphic and Lithologic Characteristics of Tsunami Deposits and an Evaluation of Possible Paleo-Tsunami Deposits in Southern California  Thesis Abstract.  Thesis Committee Ivan Colburn, Pedro Ramirez, Kim Bishop

The results of a comparison and analysis on the general sedimentary characteristics of tsunami and storm surge deposits developed from published reports on the sedimentary aspects of historical and paleo-tsunami and storm surge deposits around world indicate that both types of deposits are comprised of mainly marine silt and sand with minor amount of bioclastics such as shells and plant fragments.  Other common characteristics of tsunami and storm surge deposits are their sheet-like shapes, thickness range (from several cm to 5 m), landward thinning, clast size range (from clay to boulder), poorly sorted sediment, normal grading, better sorting upward and landward, and deposition on a sharp, irregular/erosional surface.  Because tsunami and storm surge deposits share many common sedimentary characteristics, it is difficult to differentiate them in the field.

However, the unique characteristics occurring only in tsunami deposits, such as inclusion of deep water foraminifera and association with co-seismic sedimentary features, such as dikes, sills, and volcanoes composed of sand, are reliable characteristics to differentiate tsunami deposits from storm surge deposits.  Some other features that could be used as diagnostic characteristics of tsunami deposits are their relatively large aerial extent being over 1.6 km inland and 15 km along the shore, inclusions of mud rip-up clasts, and bimodal sand grain distribution.  Also normally graded multi internal layers that may indicate bi-directional flows patterns such as imbrications of elongate clasts are common feature seen in tsunami deposits.  

On the other hand, the features that are seen in storm surge deposits but not in tsunami deposits are fossiliferous texture, grain size increasing landward, thin sand/silt laminations, a turbidite-like stratigraphic sequence including ripple marks, parallel bedding, and low angle/hummocky cross stratifications. 

All of the foregoing characteristics of tsunami/storm surge deposits have been used to evaluate four late Pleistocene stratigraphic successions along the coast of southern California.  These successions are located at 2^nd St. in San Pedro, Bolsa Chica in Huntington Beach, Superior/PCH in Newport Beach, and Crystal Cove at Crystal Cove State Park.  The late Pleistocene successions rest on a wave cut benches cut into the Miocene and into the early Pleistocene formations.  The late Pleistocene succession is mainly comprised of silt to coarse sand with gravels, and includes thin, poorly sorted, fossiliferous, gravely sand layers.  At 2^nd St. and Crystal Cove, the late Pleistocene succession is less than 100 cm thick and is very fossiliferous.  The most significant feature seen at 2^nd St. is that U- and V-shaped channels entrenched in the surface of the wave-cut bench.  The channels are filled with the poorly sorted late Pleistocene deposit and the angular shells and shell fragments in the deposit have a preferred orientation parallel to the irregular basal surface.  In contrast, the late Pleistocene succession in Crystal Cove rests on the flat surface of wave-cut bench and has parallel bedding accentuated by the horizontal preferred orientation of abraded shells and shell fragments.  At the Bolsa Chica and Superior/PCH localities, the late Pleistocene succession is approximately 25 m and 30 m thick.  In Bolsa Chica, it mainly consists of well-washed gravely sand while it is finer at the Superior/PCH site.  At both localities, the late Pleistocene succession fines upward and includes several thin fossiliferous layers and reddish stained intervals.  In Bolsa Chica, the late Pleistocene succession includes lenticular shaped channels filled up with poorly sorted sediment and cross-bedded sandstone.

The results of evaluation indicate that the late Pleistocene succession at the four localities has more sedimentary features that resemble to storm surge deposits than that they do tsunami deposits.  For example, all four localities include fossiliferous features and parallel beddings that resemble to storm surge deposits but inclusions of rip-up clast that are the characteristics of tsunami deposits are seen at only one locality.  However, possibility of tsunami deposits is not completely denied by this evaluation method.  

The late Pleistocene succession can also be totally different in origins such as beach or fluvial.  The shell rich, poorly sorted gravely sand that is seen in the late Pleistocene successions is also found in the present beach environment.  The fine-upward sequence and presence of iron hydroxide stains seen in Bolsa Chica and Superior/PCH are also common in fluvial depositional setting.  Furthermore, the stratigraphic pattern seen in the four localities selected for this study – an erosional basal surface on old marine terrace, the late Pleistocene shelly sand beds, and the non-marine cover – is observed in the broad coastal region from Northern California to all the way south to Baja California.  Such regional distribution is beyond the maximum extension distances of tsunami and storm surge events.  Therefore, the late Pleistocene succession was likely formed by normal marine coastal processes such as beach and fluvial.

Angelique C. Hamane, 2005, Texture and Phenocryst Mineralogy in a Large Dacite Dike in the Little Chuckwalla Mountains, California (Thesis Committee Members: Pedro Ramirez, Kim Bishop, David Mayo)

The Little Chuckwalla Mountains (LCM), located in Riverside County, California, expose Tertiary volcanic rocks ranging from basalt to rhyolite (Crowe et al., 1979; Mayo, 1993; Mayo and Cole, 2001).  These rocks demonstrate the widespread volcanism that was prevalent approximately 25 million years ago when the Farallon plate was being subducted beneath the North American plate (Burchfiel and Davis, 1972; Dickinson, 1981).  Swarms of dacite and rhyolite dikes crosscut andesitic lavas in the lower part of the section to feed lava and ash deposits in the upper part of the section.  The largest dacite dike is approximately 100 m thick and exposed along strike for 2,250 m.  Tilting of 30° around an axis perpendicular to the dike permits sampling of dike rock at paleodepths ranging from 0 to about 1,100 m.  Plagioclase laths in the trachytic groundmass increase in abundance and size with depth.  Common phenocrysts include plagioclase + biotite + opaques + sanidine + augite + hornblende + fayalite.  Specific phenocryst assemblage and texture vary with paleodepth.  Plagioclase, biotite, sanidine and augite are present at all paleodepths.  Fayalite is only observed in the surface lavas.    Hornblende is rare and only found at the deepest 1,100 m paleodepth.  The apparent replacement of hornblende at 1,100 m by fayalite in surface lavas records the drop in water pressure as the magma approached the surface.  Similar phase changes have been observed in experiments with H20-undersaturated granitic magma (Naney, 1983). 

Sheila K. Morrissey, 2005, Wildfire-Induced Changes to Hydrologic  Flowpaths in the San Bernardino Mountains, California (Thesis Committee: Laura Rademacher, Richard Hurst, Terri Hogue)

 The fall 2003 Old Fire caused physical and chemical changes to mountain front watersheds in the San Bernardino Mountains resulting in the formation of a hydrophobic layer, decreased transpiration, decreased infiltration, altered stream water chemistry, and increased runoff. Consequently, post fire storm-event floods in December 2003 killed sixteen people in the canyon communities of Waterman Canyon and Devore. This study of wildfire-induced changes to hydrologic flowpaths in two burned (City and Devils Creeks) and one unburned (Mill Creek) based on soil property and water chemistry changes will help to further current understanding of watershed response and recovery in the wake of wildfires. 

 Pre-fire runoff ratios from water years 2000 through 2003 were compared with the runoff ratio from post-fire water year 2004 for the burned City Creek watershed. There was less rainfall in the San Bernardino Mountains during all of the years used in the runoff ratio calculations compared with the previous ten years. Pre-fire City Creek runoff ratios averaged 0.103 with a standard deviation of 0.021. The post-fire runoff ratio for City Creek was 0.277, significantly greater than both the average and annual maximum pre-fire runoff ratios. The increase in the post-fire runoff ratio in the City Creek watershed indicates that an increased proportion of precipitation was converted directly into runoff and that overland flow was likely to have an increased influence on stream flows 

 End Member Mixing Analysis (EMMA) was performed using geochemical data from the two burned and one unburned watersheds. The EMMA model created for the San Bernardino watersheds utilized three end-members with unique solute concentrations for groundwater, soil water, and precipitation plus overland flow. Using three end-members rather than two (“new” and “old” water) as in a two component hydrograph separation provides a means for evaluating changes in flowpaths as the watershed recovers, as soil water and groundwater are likely to have very different residence times within the watershed. Changes in the relative proportions of the end-members which mix to form stream water were calculated using EMMA in the burned City Creek watershed and the unburned Mill Creek watershed at Monkeyface Falls. Results suggest an increase in the proportion of direct precipitation to the stream channel and short residence time overland flow component of stream water in the burned watershed. However, the exact proportions could not be determined because stream water concentration data from pre- and post-fire samples were not entirely bounded by the end-members used for this model. 

 The increase in the proportion of precipitation and overland flow is may have resulted from the formation of a hydrophobic layer in the burned soils. Hydrophobicity of burned soils was measured using water drop penetration tests (WDPT). Seven months after the Old Fire, burned soils in the formerly chaparral portion of the City Creek watershed were non-repellent to moderately repellant, with a maximum infiltration time of 120 seconds. Nineteen months after the Old Fire, maximum infiltration time in City Creek soils was only 18 seconds. The soil hydrophobicity created by chemical processes during the wildfire likely led to the increase in precipitation and overland flow stream components. Relatively dilute precipitation decreased the chloride concentration in the stream water after the fire. Additionally, the chloride mass balance of the watershed suggests the decrease in vegetation caused a decrease in evapotranspiration, which normally concentrates chemical solutes in soil water. A chloride mass balance of pre- and post-fire stream water from Devil’s Canyon indicated a 75% loss of evapotranspiration after the Old Fire.  Contrary to the decrease in stream water chloride, a study of the chemical changes in weathering derived solutes in streams from the pre- and post-fire City Creek watershed and the unburned Mill Creek watershed indicated a post-fire increase in net export of calcium, magnesium, potassium, sodium and sulfate. The increase in concentration of these solutes in stream water is probably due to increased weathering during flooding and decreased vegetative uptake. Net export of nitrate, however, decreased post-fire. The decrease in nitrate export was likely due to volatilization of nitrate during the wildfire.