4

Origin and Evolution of Local Off-Lake Sources in Owens Valley

As described in Chapter 3, a number of sources of PM₁₀ exist in the Owens Valley Planning Area (OVPA) that are outside the regulatory shoreline of Owens Lake, including flood deposits, Keeler Dunes, Olancha Dunes, sand sheets, alluvial fans, up-valley sources, and a variety of anthropogenic disturbances. In this chapter, the panel evaluates the origin and evolution of these major off-lake sources, or those that might become important sources in the future. Some anthropogenic disturbances that were identified as causing exceedances in Chapter 3 (e.g., landfill, roads, construction) are not expanded upon in this chapter as they are outside the scope of this study (Table 1-1).

There are several methods for determining the origin of emissive landforms and how they developed over time. Radiocarbon dating can provide ages of vegetation, flooding, or evidence of past environments, such as immersion in water or burial of soils. Luminescence dating utilizes sand-sized grains and the property they contain that bleaches to zero as the grain is exposed to sunlight on a transport path. This method can provide the best absolute dating evidence for the timing of when the grain is redeposited in a landform and covered, thus providing information on the history of fine-grained sediment mobility on the landscape. Geochemical and grain size information can also be used to trace sedimentary sources and processes. Finally, aerial imagery and photography can document geomorphic and vegetative changes to landscapes. Geologists and geomorphologists typically use these multiple lines of evidence to infer the origins and evolution of landforms, including the potentially dust-producing ones discussed here.

WINNOWING HYPOTHESIS FOR OFF-LAKE DUST

A process that is potentially relevant for the origin and evolution of a number of PM₁₀ sources in the OVPA is “winnowing.” As described in Chapter 3, the term “winnowing” refers to the removal of dust from areas above the regulatory shoreline, which had historically originated on the lakebed. This process was described by Ono and Howard (2016) and was included as Appendix G in the 2016 State Implementation Plan (GBUAPCD 2016). Ono and Howard (2016) found that there was a correlation between on- and off-lake exceedances at the Dirty Socks monitor, and that both exceedances were decreasing in frequency and in average concentrations over time. Their analysis implied that the bulk of PM₁₀ emission from off-lake sources was originally derived from on-lake sources, and that as PM₁₀ emissions from on-lake sources decreased, the derivative PM₁₀ in off-lake sources would also decrease. The report states that:

More than 8 years later, it appears that the relationship between on- and off-lake dust sources does not hold. Analysis of exceedances from 2001 to 2023 based on consistent metrics indicates that, contrary to the expectations set out in Ono and Howard (2016), the number of off-lake exceedances do not show a significant downward trend over time, even as on-lake exceedances emissions have continued to decrease (Figures 3-6, 3-9, 3-12, 3-18, 3-22, and 3-24). The panel’s analysis of estimated emissions at the two sites with long-term data showed similar results for Dirty Socks (Figure 3-27; see also Appendix A). The panel did observe a significant downward trend in estimated off-site PM₁₀ emissions at Keeler (Figure 3-28), which is to be expected in response to the Keeler Dunes Dust Control Project (see Box 6-1). Most current dust emissions from off-lake sources are likely not a result of resuspension of material that originated on the lakebed. Instead, the presence and frequent replenishment of highly emissive flood deposits provides ample fine particulates that can be emitted as dust as long as the horizontal flux of sand-sized particles in saltation is sufficient to emit dust-sized particles from the surface. Thus, the panel finds the premise of Ono and Howard (2016) that there is a “limited supply of sand and dust in these off-lake areas” to be false. Given this and the lack of correlation between on- and off-lake exceedances and estimated emissions, the panel finds that winnowing will play a minimal role in reduction in future off-lake PM₁₀ exceedances.

NORTHEAST SIDE OF THE LAKE

The northeastern side of Owens Lake is host to several dune fields, sand sheets, and alluvial fan complexes that deliver and rework sediments from the neighboring Inyo Mountains, pre-existing aeolian deposits, former floodplain and delta deposits of the Owens River, and from the Owens Lake bed itself. Some of these landforms, like the Lizard Tail Dunes and Swansea Dunes, have existed on the landscape for long periods of time without substantial change (Lancaster and Bacon 2012; Lancaster et al. 2015). The Lizard Tail and Swansea dunes have not been tied to many exceedances (Chapter 3). In contrast, other landforms to the northeast of the lake, like the Keeler Dunes, have overgone major changes over the last century and are significant contributors to exceedances (Chapter 3).

Keeler Dunes

The Keeler Dune field consists of sand deposits overlying late Holocene alluvial fan deposits that are located below and to the west of the Slate Canyon/Keeler Alluvial Fan Complex (Figure 4-1). The Keeler Dunes and nearby alluvial fan complex are large continuing sources of off-lake exceedances in the vicinity of Owens Lake (Chapter 3). In 2014, the Great Basin Unified Air Pollution Control District (or District) started a project to reduce aeolian transport and PM₁₀ emissions on the Keeler Dunes (the Keeler Dunes Dust Control Project), which involved reducing sand transport using artificial roughness and introduction of plant species in an effort to revegetate the dunes (see also Box 6-1).

The District and the Los Angeles Department of Water and Power (LADWP) have provided competing explanations for the activation of the dune field and resultant production of PM₁₀. Here, the panel outlines the contextual elements common to both views, details its understanding of the two explanations for activation of the Keeler Dunes, and evaluates the evidence for each.

There is evidence that the Keeler Dunes were a largely stabilized dune field covered mainly by black greasewood (Sarcobatus vermiculatus) at the beginning of the 20th century (Figure 4-2). This dune field probably looked very much like the more northerly Swansea and Lizard Tail dunes and like the southern portion of the Keeler Dunes today. Researchers agree that some event or events caused destabilization of portions of the Keeler Dunes (Figure 4-3). Neither the Swansea and Lizard Tail dunes nor the southern portion of the Keeler Dunes appear to

have been impacted in the same way. The main arguments for the remobilization of the Keeler Dunes are referred to below as the “hydrological” and “aeolian” arguments for the sake of simplicity and are outlined briefly here:

1. Hydrological argument: Berms installed in 1954 and 1967 to protect Highway 136 reduced the delivery of surface runoff water to the distal portion of the Keeler/Slate Canyon Fan. As a result, the greasewood that stabilized the dunes was reduced. This line of argument was made primarily in Richards et al. (2022).
2. Aeolian argument: Diversion of water from the lake and subsequent desiccation of the lakebed exposed significant amounts of sand that was blown into the Keeler Dune field. The increase in aeolian transport overwhelmed the vegetation’s ability to stabilize the dunes due to feedbacks between the increased sand flux (and abrasion potential) and declining vegetation cover. This line of argument was made primarily in Lancaster and McCarley-Holder (2013).

These two arguments will be evaluated more below, but first, the common geomorphic and hydrologic context for the Keeler Dunes will be discussed.

Geomorphic Context

Sands from the Keeler Dunes are medium- to fine-grained, moderately to poorly sorted, and have a quartz content of 37–38 percent and a total feldspar content of 40 percent (Lancaster et al. 2015); (Figure 4-4). The composition and mineralogical maturity of the sands from the Owens River and the Keeler Dunes are similar, although the dune sands are slightly more quartz rich, which often occurs in aeolian settings as other minerals (e.g., feldspars) are abraded during aeolian transport, or weathered over time (Lancaster et al. 2015; Muhs 2004). In contrast, sands on the Slate Canyon/Keeler Alluvial Fan Complex have distinctly different mineralogy with higher calcite and lower feldspar content (Figure 4-4A; Lancaster et al. 2015).

Regardless of when sand was deposited (i.e., pre- or post-diversion of water from the lake), mineralogical composition and grain size distributions indicate that the primary source of sand for dune fields in this region appears to be sediment from the Owens River, derived initially from the physical breakdown of granitic rocks of the Sierra Nevada that was then remobilized by wind from the lakebed and exposed river delta into the dune fields during times of low lake levels (Lancaster and Bacon 2012; Lancaster and McCarley-Holder 2013; Lancaster et al. 2015). Past drops in lake levels and periodic floods are known to have increased sand supply to the Owens Lake bed and subsequently to dunes proximal to the lake (Bacon et al. 2018; Lancaster and McCarley-Holder 2013; Lancaster et al. 2015). In recent geological time, optically stimulated luminescence (OSL) ages from the Lizard Tail Dunes indicate two periods of aeolian sand accumulation: 1192–1302 CE and 1592–1712 CE. These periods are also represented in the Keeler Dunes area, suggesting extensive aeolian sand accumulation around 1282–1412 CE and 1563–1712 CE (Bacon et al. 2018; Lancaster and Bacon 2012). The older of these periods closely follows regression from the 3,618 ft (1,103 m) lake high stand that occurred around 1112–1282 CE (Bacon et al. 2018; Lancaster and Bacon 2012), which would have resulted in exposure of lake plain sediments as lake levels lowered. Lancaster and Bacon (2012) provided luminescence ages that indicate aeolian sand was also reactivated during 1930–1976.

Flood deposits have a much higher potential for PM₁₀ emissions compared to aeolian sands and interfluve¹ fan deposits (Kolesar et al. 2022b). Flood deposits typically occur in distal areas of the fan complex fed by recently active channels and/or within low-lying interdune basins where the dunes locally dam flood flows. In the Keeler Dunes, fine textured (clay–silt) “flood silts” exist within interdune areas as depicted in Figure 4-5 (Lancaster and Bacon 2012) and are generally interbedded with aeolian sands or deposited as a cap on underlying aeolian sands. Periodic flooding is a natural occurrence in desert landscapes that results in mobilization of sediments from slopes, channels, ephemeral washes, and alluvial fans within watersheds. Flood flows also interact with and rework other

___________________

¹ Interfluve is the “relatively undissected upland between adjacent streams flowing the same general direction” (American Geological Institute 1983).

sedimentary deposits, including aeolian dunes and sand sheets. As such, flood deposits are an episodically renewing part of the natural mosaic of sedimentary features, including at Keeler Dunes.

Hydrologic Context for Vegetation Around Owens Lake

As outlined in chapter 2, there is considerable evidence at present of near-surface groundwater at the margin of the 3,600-ft regulatory shoreline. The wetlands, springs, and seeps that ring the lake (Figure 2-7) and other evidence (Meyers et al. 2021) indicate remarkable stability of groundwater supply in the vicinity of Owens Lake that has not been impacted by the diversion of Owens River, nor by groundwater pumping up-valley. Modeling by Richards et al. (2022) suggests that depth to groundwater in the Keeler Dunes complex is 0–9 m.

Greasewood is a phreatophytic (groundwater-dependent) shrub that stabilizes dunes in the northeastern side of Owens Lake, including around Keeler (Figure 4-2). Studies have noted the capacity of greasewood (Chimner and Cooper 2004; Nichols 1994) to optimize the uptake of water from shallow or deep water sources based on prevailing conditions and to switch to groundwater when vadose zone soil moisture declines (Devitt and Bird 2016). Studies have shown that greasewood can use groundwater to depths of 33–43 ft (10–13 m; Chimner and Cooper 2004; Devitt and Bird 2016; Garcia et al. 2015; Nichols 1994). In the Mono Lake area and the vegetated dune areas at Owens Lake, greasewood has been shown to have roots down at least 10–16 ft (3–5 m) and 15–20 ft (4.5–6 m), respectively (Donovan, Richards, and Muller 1996; Mike Aspinwall, Formation, personal communication, July 30, 2024). Elmore, Mustard, and Manning (2003) reported that the average depth to groundwater in greasewood communities in Owens Valley is 11 ft (3.4 m).

Evaluation of the Hydrological Argument for Keeler Dune Destabilization

Richards et al. (2022) state that “construction of flood control berms (1954, 1967) cut off surface overflow events on the alluvial fan, resulting in subsequent very low plant cover and significant sand movement.” Richards et al. (2022) summarizes the evidence in support of the Hydrological Argument for Keeler Dune destabilization, and begins its vegetation analysis in 1944, when the first aerial imagery from the area is available. However, by 1944, considerable time had passed since the final exposure of the nearby lakebed (the North Sand Sheet was exposed between 1917 and 1920, Figure 4-6). Thus, by 1944, sand from the North Sand Sheet (Lancaster and McCarley-Holder 2013) could have already impacted dune stability.

The argument regarding the cause of changes in vegetation cover (Richards et al. 2022) hinges on disruption of periodic overland flow. With hydrologic modeling, Richards et al. (2022) showed that for some years, the amount of surface flow events in areas below the berm would have greatly decreased. Richards et al. (2022) hypothesizes “that interruption of the surface flows by berm construction would have greater impacts on upland, alluvial fan vegetation than on the groundwater-dependent vegetation of the dune complex (Perkins et al. 2018).” This is consistent with other studies that show that shallow-rooted, rainfall-dependent vegetation can be affected from upstream berms (Schlesinger and Jones 1984). The areas immediately below the berms (e.g., regions B3 and C4 in Figure 4-7A), would have been the most affected as these are mostly dominated by precipitation-dependent “upland vegetation” (i.e., shallow-rooted non-phreatophytic vegetation such as most Atriplex spp., saltbush, including A. confertifolia, A. parryi, and A. hymenelytra). In fact, a difference in vegetation cover was clear above and below the berm when the panel visited in May 2024 (Figure 4-8). However, Figure 4-9 from Richards et al. (2022), which compares the vegetation cover in 1944 and 1996, does not show large changes in vegetation distribution in the area immediately below the berm in region B3 despite the construction of the berm after 1944. This evidence suggests that there may be other processes that are more strongly controlling vegetation cover. However, in general, the panel finds that Richards et al. (2022) is successful in showing that areas of the Keeler/Slate Canyon Fan that are not groundwater-dependent have higher cover of “upland” (saltbush) vegetation in the absence of highway berms compared to a case in which the berms are present.

In contrast, Richards et al. (2022) does not clearly show that areas that would have had consistent shallow depths to groundwater (e.g., A3 and the lower portions of B3 in Figure 4-7A) were impacted by the berm. Their analysis is consistent with A3 and the lower reaches of B3 having relatively high (approximately 30–40 percent)

groundwater-dependent vegetation cover given modeled groundwater depths (similar to B4 and B5, see Figure 4-7A) in the absence of any “non-hydrologic” impacts, such as aeolian transport. Indeed, Richards et al. (2022) argues that area A3 (Figure 4-7A) has low vegetation cover compared to other areas with relatively shallow groundwater, possibly due to high sand flux in the area. These impacts to vegetation appear to be present in 1944 and therefore predate any hydrological impacts of the berms themselves. This argument is thus entirely consistent with the view that the northern Keeler Dunes were previously destabilized from some other means (e.g., transport of sands from the North Sand Sheet following water diversion from the lake, per the aeolian argument, see below). Indeed, if disruption of surface hydrology by the berms had impacted dune vegetation, it is logical for this effect to be seen most strongly where the groundwater is deepest and only later witnessed, if ever, in areas with shallow groundwater. In 1944, however, areas with shallow groundwater (A3, particularly in the north) appear to be devoid of vegetation whereas denuded dune vegetation still exists in areas with deeper groundwater (such as in the lower portion of B3; Figure 4-7A). This pattern is thus inconsistent with a hydrologic explanation for vegetation loss.

It appears that the modeling approach used by Richards et al. (2022) has little ability to inform the effect of the berms on the ecohydrology of groundwater-dependent vegetation. For example, the modeling approach does not allow the berm to have any effect on the groundwater table, while plant transpiration is prescribed by unchanging root density distributions that depend only upon prescribed vegetation type (upland vs. groundwater-dependent) with root density mainly in the vadose zone. Thus, vegetation growth for groundwater-dependent vegetation, in the model context, always depends on vadose-zone moisture in a strictly prescribed way, even in areas where vegetation has access to groundwater. This approach also does not account for adaptations that vegetation may undergo as a result of changing ecohydrologic conditions, such as root pruning in the vadose zone and increasing water uptake in the saturated zone, perhaps through growth of new roots (e.g., Chimner and Cooper 2004; Nichols 1994). The panel notes that the type of modeling done by Richards et al. (2022) can provide important insights, but the details and limitations of such models are important to consider. Absence of any adaptations to plants in

response to changing vadose zone water availability or groundwater depth make it impossible to infer from the modeling what impact berm-related changes might have on vegetation in a realistic sense.²

Although this study and others show that berms that disrupt overland flow can impact shallow-rooted, rainfall-dependent vegetation (Richards et al. 2022; Schlesinger and Jones 1984), it seems unlikely that disruption of overland flow from the berms in Owens Valley had a major impact on greasewood’s ability to access groundwater. Instead, to maintain productivity during dry periods, including after creation of the Highway 136 berms above Keeler Dunes disrupted overland flow, greasewood should be able to access groundwater (or the capillary fringe, which reaches above the groundwater table about 2 ft [0.6 m] for sands and 3 ft [>1 m] for finer soils [Lu and Likos 2004]). Data presented by Mike Aspinwall to the panel on July 20, 2024, indicates that greasewood in the Keeler Dunes area can access and take up groundwater throughout the year (Box 4-1).

Furthermore, greasewood at other sites around Owens Lake do not appear to rely on significant overland flow to maintain productivity. For example, some portions of the Swansea Dunes maintain greasewood cover but do not sit at the distal edge of a large fan like the Slate Canyon/Keeler Fan. In fact, the distance upslope from the Keeler vegetated dune area to the Highway 136 berms is greater than or equal to the distance from the southern portions of the Swansea Dunes to the nearest mountain front, suggesting that greasewood in the area does not rely on significant overland flow to maintain productivity. Evidence from Swansea Dunes, therefore, suggests that greasewood vegetation can be sustained by less moisture from overland flow than is likely received by the Keeler Dunes. Therefore, disruption of overland flow from highway berms above Keeler Dunes would be unlikely to cause the death of mature individuals or cause vegetation collapse.³

In summary, the panel evaluated the main paper used to support the hydrological argument for Keeler Dune destabilization (Richards et al. 2022). Although this paper appears to support reduced vegetation cover in upland areas, there is little support for the idea that highway berms led to hydrologic changes that destabilized groundwater-dependent vegetation of the Keeler Dunes. Additional evidence, such as apparent continued utilization of groundwater by greasewood in the area of Keeler Dunes and continued stability of Swansea Dunes, despite limited access to runoff due to its topographic position, further suggests that destabilization of the Keeler Dunes was not caused by hydrologic changes due to installation of highway berms in the 1950s and 1960s.

___________________

² This paragraph was edited after release of the report to clarify the modeling approach used by Richards et al. 2022.

³ This paragraph was edited after release of the report to clarify area of comparison.

Evaluation of the Aeolian Argument for Keeler Dune Destabilization

An alternative explanation for the destabilization of a vegetated Keeler dune field has been offered to the panel. In what we are calling the aeolian argument, desiccation of the lake exposed expansive areas of sand stored in the Owens River delta and North Sand Sheet (an area of the Owens Lake bed immediately south of the Owens River delta; Figure 4-10). Aeolian processes then transported this lakebed material into the area of the Keeler Dunes (particularly the northern portion of the dune field). At the beginning of the 20th century, this portion of the dune field was largely stabilized, and although bare interplant areas may have experienced some aeolian transport, the volume of sand moving through the system was insufficient to cause large-scale mortality of stabilizing vegetation. According to this argument, the addition of sand from the lakebed after exposure of the proximal North Sand Sheet by 1920 (Figure 4-6), however, was sufficient to result in vegetation loss on the formerly stabilized dunes. The initial reduction of plant cover led to increased aeolian activity through the combined effects of reduced sheltering by vegetation as well as increased sediment supply, which generated a positive (amplifying) feedback of vegetation loss and large-scale remobilization of the northern part of the dune field. Lancaster and McCarley-Holder (2013) tracked the evolution of the dune field from 1944 to the beginning of the 21st century, and Lancaster and Bacon (2012) provide additional chronological and stratigraphic context.

Photographic evidence from the 1910s (Figure 4-2) suggests portions of the Keeler Dune field was stabilized by vegetation before diversion of water from the lake exposed considerable sediments available for aeolian

transportation. After exposure of the North Sand Sheet on the lakebed, photographic evidence further suggests that there was significant aeolian transport in the Keeler Dunes area, which impacted Highway 136 enough to necessitate installation of sand drift fences. Furthermore, train records indicate that blowing sand and sand drifts were a problem for the railroad in 1954 between milepost 573 and 575, which corresponds approximately to the area of active dunes in the 1944 aerial photo (Figure 4-11; Grace Holder, personal communication, May 2024).

Thus, historical records appear to support the aeolian argument and strongly suggest that the destabilization of the vegetated Keeler Dunes had begun considerably before the building of the protective berms in 1954 and 1967. The OSL age of 1930 and its associated 1 sigma uncertainty range of ± 25 years (Lancaster and Bacon 2012), which is common for relatively young deposits, presents the possibility that the aeolian sand deposit from which the OSL sample was collected was emplaced sometime between 1905 and 1955. This age range overlaps both with the diversion of water and subsequent desiccation of the lakebed as well as the emplacement of the berms in 1954. By itself, this date does not prove destabilization of the Keeler Dunes before 1954, though it is consistent with this view. Without additional sampling, existing OSL data are not conclusive concerning when periods of major dune reworking occurred. Using augers to complete an OSL coring campaign across the sand dunes/sand sheets could provide a more comprehensive set of young ages that are reflective of recent dune activation. While quartz would be the primary target of these studies (as the dominant mineral and best suited to produce higher-resolution OSL ages), feldspar may also provide helpful OSL ages and distinctive luminescence characteristics that allow for increased confidence in the OSL chronology.

Lancaster and McCarley-Holder (2013) note that since 1944, when the first aerial imagery exists, “the dunefield has undergone significant changes, including development of well-defined linear and crescentic dunes from an initial small area of partially vegetated dunes, resulting in an increase in the area of the dunes by a factor of 3 since 1944” (Figure 4-12). This has resulted in a relatively large area of active aeolian transport and surface deflation, which has the potential to produce PM₁₀ emissions, especially with the presence of interbedded flood silts and aeolian sands within the Keeler Dunes area (Kolesar et al. 2022b; Lancaster and Bacon 2012), which serve as a source for PM₁₀ material.

While the southern Keeler Dunes continued to develop and expand from the 1980s through early 2000s, significant wind erosion occurred on the upwind (northern) margins of the dune field (Lancaster and Bacon 2012) with as much as 4 ft (1.2 m) of surface deflation occurring in the interbedded flood silts and aeolian sands. This upwind erosional response corresponds with implementation of dust control measures in 2000, which has served to starve the Keeler Dunes of sand supply and transitioned the system into a negative sediment budget (Lancaster and McCarley-Holder 2013). Meanwhile, the southern region of the Keeler Dunes continued to expand and migrate southeast by reworking existing dune sands. As such, it is apparent that the draining of Owens Lake in the early 20th century consequently increased sand supply to and expansion of the Keeler Dunes, which remain an active dust emissions source. In turn, this has complicated the patterns of fine-grained flood deposits and the extent of their deposition from the Slate Canyon alluvial fan/Keeler Fan Complex within the dunes that provide additional sources of PM_10.

Several papers have been published that disagree with aspects of this overall narrative. These will be discussed here:

1. Blanton, Kolesar, and Jaffe (2022) have argued that the Keeler/Slate Canyon fan is likely the source of sediment for the Keeler Dunes, at least since construction of the protective berms. However, hydrologic and hydraulic modeling cannot be substituted for mineralogical and geochemical analysis in determining sediment provenance. Such arguments are unconvincing in light of the mineralogical evidence provided by Lancaster et al. (2015; Figure 4-4), which indicate that the mineralogical composition of the dunes is very similar to that of the Owens River and quite dissimilar to the mineralogy of the Keeler/Slate Canyon Fan. More geochemical analyses could further assess this line of evidence.
2. Schmid et al. (2022) argued that dune volume within the dune field was relatively stable over the period of active disturbance (after 1944), which is inconsistent with delivery of large amounts of sand from the Owens Lake bed that could destabilize the Keeler Dunes. The panel has identified several uncertainties in the Schmid et al. (2022) analysis. First, Schmid et al. (2022) only starts its analysis in 1944, at least

1. 20 years after the exposure of the nearby lakebed and thus the initiation of transport from the lakebed to the Keeler Dunes. While the panel recognizes this analysis is limited by the availability of the first available aerial imagery, sufficient delivery and storage of sediment in the Keeler Dunes could have occurred before 1944, the effects of which could have been experienced for decades to come. Second, this argument does not consider potential losses of sand from the dune field through transport in the thin sand sheet that is clearly progressing up the Slate Canyon/Keeler Alluvial Fan Complex (Figure 4-13). Even if the volume estimates of Schmid et al. (2022) are accurate, addition of sand from the Owens Lake bed from the west could have been offset by loss of sand in the thin but extensive progressive fan sand sheet. Third, the volume estimates of Schmid et al. (2022) are of insufficient precision to support conclusions about Keeler Dune volumes. The method used to estimate volumes from digital elevation models (DEMs) generated from historical aerial photography in this study is somewhat unconventional. The approach of Schmid et

2. al. (2022) is subject to notable uncertainty resulting from the resolution and overlap of historical imagery when used to generate DEMs using traditional photogrammetry. Additional uncertainty is added when defining and interpolating a “sub-dune surface” plain for each photo year, which is used to derive dune heights and raster volumes. More conventional methods use spatial statistics to identify significant elevation changes between successive DEM intervals (Wheaton et al. 2010) and calculate volumetric changes only on locations of significant differences in elevation between the DEMs, not referenced to some arbitrary underlying surface. These more conventional approaches identify multiple error sources, compound them, and remove insignificant raster values (i.e., those that fall below the detection threshold defined by the compounded error budget). Thus, it is not clear that the approach of Schmid et al. (2022) is sufficient for accurate and precise calculation of dune field volumes, and this argument about dune field volumes does not consider additions to the Keeler Dunes prior to the first available aerial imagery (1944) or losses of sand through upslope transport in a thin sand sheet.
3. Kolesar et al. (2022a) estimated a novel index of sand transport concluding that sand transport on the lake occurred in a direction that does not support transport of sediment from the lakebed onto the area of the Keeler Dunes. This study suggests that the wind regime in the area does not support the hypothesis of additional sediment to the Keeler Dunes sourced from the dry lakebed. The study used wind records from a single year (2001) and wind direction measurements from stations on the lake (A Tower and Delta), which are a considerable distance from the historical shoreline in the vicinity of Keeler (approximately 3–8 km). The panel has identified considerable uncertainties associated with this study. First, 1 year of wind data is too minimal to determine decadal-scale patterns in transport direction. Second, as pointed out in Holder et al. (2024), the wind screening conducted by Kolesar et al. (2022a) fails to recognize that “the surface crusting,” which is the justification for the screening, “generally occurs on clay and silt dominated soils and is not prevalent on the thick sandy deposits found on most of the northern portion of the bed of Owens Lake (termed the North Sand Sheet) where the analysis was conducted.” As a result, the longer records used in the estimates of net sand transport direction (105 degrees, ranging from 17 degrees to 145 degrees) produced by Ono et al. (2011) and Lancaster and McCarley-Holder (2013) are less uncertain (Figure 4-14). These directions support transport from the North Sand Sheet to the area of the Keeler Dunes (Figure 4-14). Third, it is also important to note that resultant drift potential or net transport directions, when calculated from station data, do not fully represent sediment dynamics in areas with complex topography or variable vegetation. On a smooth, unvegetated surface, winds from various directions may move a sand grain in the direction of net transport over a period of time. However, in complex terrain or areas with variable vegetation cover, sediment cannot be expected to experience the same transport patterns (Figures 4-15 and 4-16). Use of winds at a single point for calculation of resultant sand drift potential (Kolesar et al. 2022b; Lancaster and McCarley-Holder 2013) will, therefore, not capture actual sediment movement over areas with complex topography or variable vegetation cover. Indeed, the arrangement of topography in the vicinity of Keeler Dunes is likely to steer transport north of the generally east/southeast resultant drift potentials predicted from winds measured on the lakebed (Figure 4-15). And because vegetation above the 3,600-ft shoreline protects sand blown into it from being blown back out again (per Figure 4-16), transport off the lake could be increased farther still from what might be expected from resultant sand drift potential estimated from winds measured on the lakebed. Finally, the analysis of Kolesar et al. (2022b) and consequent arguments do not account for a wealth of geomorphic information that is probably more relevant for understanding decadal-scale transport than a single year’s data. For instance, the 1944 aerial photos show north-facing slip faces on relatively large dunes on the lakebed (Figure 4-7) indicating significant periods of dominantly northward wind transport near the 3,600-ft contour, which likely would have been able to transport sediment off the lake in the area of the Keeler Dunes. Once transported into the Keeler Dunes, microtopography of the dunes and/or dune vegetation could prevent transport from opposing winds back onto the lakebed. In addition, images of the area tend to corroborate the more northward aeolian sand drift potential roses of Lancaster et al. (2015) compared to those of Kolesar et al. (2022b). Figure 4-13 shows lighter toned, finer textured aeolian sediments on the fan complex. The morphology and orientation of these deposits support transport in an

1. upslope direction from the general direction of the lakebed as suggested by the resultant sand drift potential vector. The darker toned surfaces to the right of the line in Figure 4-13 reflect coarser sediments of the alluvial fan/wash complex. The boundary between aeolian and alluvial surficial sediments is dynamic and has generally progressed up slope in response to continued sediment supply from the dunes. This movement is evident by comparing the position of the boundary between 1944 and 1993. The migration continues to present day (Figure 4-13D), despite implementation of dust mitigation efforts on the lakebed in the early 2000s. Small dunes and aeolian sand sheets are also found upwind (southwest) of, and migrating over, the flood diversion berms built in 1954 and 1967 on the northeast side of Highway 136 (Holder et al. 2024). The boundary between lighter- and darker-toned sediments appears to be influenced

1. by periodic floods on the fan complex that transport some of the aeolian sands back toward the dunes and lower valley floor, as evident by comparison of the 1993 to 2014 images (Figure 4-13B, C) or the image that overlays data from 1985–2020 (Figure 4-13E). This structure now diverts floodwaters and fine sediments from the fan complex to two focal locations approximately 1.3 km apart along Highway 136. These locations require regular maintenance following flood events to clear the highway and drainage culverts of sediments, which include aeolian sands that continue to migrate back upslope from the Keeler Dunes and lower elevation areas.

Panel Assessment of the Origin and Evolution of Keeler Dunes

There is considerable historical evidence that supports the aeolian argument for the destabilization and growth of the Keeler Dunes. The counterarguments offered by Blanton, Kolesar, and Jaffe (2022), Kolesar et al. (2022a), and Schmid et al. (2022), do not present convincing refutations of the basic processes underlying the aeolian argument. In contrast, the panel found the hydrological argument to be unconvincing and inconsistent with other observational evidence. The panel finds the aeolian argument to be the most likely explanation for the current active Keeler dune field and support the view that desiccation of Owens Lake is ultimately responsible for the destabilization of the historically vegetated Keeler Dunes.

With the destabilization of the vegetated Keeler Dunes and its emergence as an active dune field, there is clearly sufficient saltation to drive PM₁₀ emissions from the flood silts in the area (as seen by the continued exceedances). Because fine-grained material for PM₁₀ emissions from Keeler Dunes are continually replenished via flood deposits, there is no reason to believe that PM₁₀ exceedances from the Keeler Dunes will stop as on-lake sediment sources are controlled. However, even though the PM₁₀ emitted from the Keeler Dunes may not have originated on the lakebed, the desiccation of the lake still increased dust emission from the Keeler Dunes by destabilizing the vegetated dunes at Keeler and increasing the horizontal flux of saltating material that, through sandblasting, emits PM₁₀ from the flood deposits. If the vegetated dunes had not been destabilized by excess sand from the dry lakebed, the current and past levels of dust emissions at Keeler Dunes would not have occurred.

Berm-Related Channel/Flood Deposits at Keeler Dunes

Several constructed berms northeast of Highway 136 intentionally modified the surface hydrology of overland flow to focus floodwaters to specific points of discharge along the highways. The panel did not examine each berm-related flood deposit individually but instead focused only on the berms installed in 1954 and 1967 on the Keeler/Slate Canyon Fan. As discussed above, the panel did not find evidence that the influence of the berms on the distal portion of the fan was sufficient to have modified the morphodynamics of the Keeler Dunes (i.e., per the hydrological argument). It is evident, however, that the berm has recently had appreciable, localized impacts on the distribution of flood deposits in the Keeler Dunes region. For instance, the 2022 aerial photograph in Figure 4-13D shows pronounced channel incision and flood channel deposits below the berms following impacts from the remnants of Hurricane Kay in September of 2022. These types of events and their related deposits appear to be infrequent as they are not evident in other aerial photographs. However, these infrequent events could still contribute appreciably to source materials for PM₁₀ emissions detected at the Keeler monitoring station. Further investigation is needed to determine how the changes in the distribution of the flood deposits due to the berms at this and other locations have changed dust emission potential.

SOUTHERN SIDE OF THE LAKE

The southern side of Owens Lake is host to another major dune field, sand sheets, and multiple alluvial channel/wash systems that deliver and rework sediments from the neighboring Coso Range. As discussed in Chapter 3, several of these features have been identified by the District as contributing sources to exceedances.

Centennial and Coso Washes

The Centennial Wash drains the northern portion of the Coso Range ending in a large alluvial fan that extends into the Owens Lake bed, and the Coso Wash similarly terminates in an alluvial fan located to the southwest of the Centennial Wash (Figure 3-11). As with all such desert washes, the Centennial and Coso wash/fan complexes consist of a series of channels set into the fan, filled with generally fine-grained (compared to the fan), wind-erodible sediment, and kept free of vegetation through occasional flows of water. Like many if not most channels, the ample presence of sand and low vegetation cover create conditions ripe for frequent aeolian transport. Additionally, since any event that brings water and sand down the channels also tends to bring PM₁₀ producing material, the PM₁₀ emission potential of these channels can be quite high. It is therefore not surprising that PM₁₀ emissions have been observed from the wash/fan complexes, especially after recent flooding caused by heavy rains from atmospheric rivers and tropical storm events (Figure 4-17; see also discussion of Shell Cut monitoring site in Chapter 3). Although flooding at Centennial Wash in 2022 and 2023 (Figure 4-18) generated more dispersed flood deposits than at Coso Wash, analysis of pollution rose data and other data in Chapter 3 suggest that Coso Wash was the dominant cause of elevated PM₁₀ at the Shell Cut monitoring on the southern side of the lake. Therefore, the committee more closely examined the Coso Wash area.

Such flood events have little to do with anthropogenic activities. However, to the extent that climate change affects the frequency and intensity of these large, rare events, the emissions that they cause may be expected to increase or decrease. The panel’s analysis of aerial and satellite imagery of the Coso Wash (Figure 4-17), shows little in the way of anthropogenic impacts that might lead to increased PM₁₀ emissions from the fan. No evidence was found, for example, of transport of sand from the exposed lakebed onto the fan, where it may have either changed the morphology of the fan/channel or where it may have provided excess saltators in the presence of PM₁₀ material. Although some highway berms or ditches were constructed in the past to reduce flood impacts to Highway 190, analysis of aerial imagery in Figure 4-17 does not suggest that any structures have altered flood flows in the Coso Wash in ways that would be likely to increase PM₁₀ emissions.

Dirty Socks Area

No evidence was found in the Dirty Socks area (Figure 4-19) of impacts that may have increased PM₁₀ emissions related to the drainage of Owens Lake. However, the panel did note that there is an area of flood deposits sitting behind a beach ridge just south of the Dirty Socks monitor. In 1944, this appeared as a relatively small area with Highway 190 cutting across one side. By 1977, this highway had been moved to the higher beach ridge to the north. From 1977 onward, the area of flood deposits behind the beach ridge has continued to grow, representing successive flood events that deposited material behind the beach. These deposits are likely possible sources for exceedances observed at the Dirty Socks station, which is immediately to the north. Although these images are not definitive, there is a possibility that moving the highway increased the potential of this small depression to store impounded water and sediment. In doing so, it may have inadvertently increased the potential of this area to contribute to PM₁₀ emissions. Detailed high resolution imagery and coring for geochronology and geochemistry could further constrain the impact the highway had on the flood deposit. In particular, coring for ages of sediment deposition using OSL could establish the cycles of flooding, the amount of sediment tied to the flood events, and fine-grained storage potential of the Dirty Socks location. This sampling would be most effective if it occurred in conjunction with a well-defined grain size and geochemical analyses of each stratigraphic unit.

Olancha Dunes

The Olancha Dunes overlie late Pleistocene and Holocene alluvial fan and lake deposits associated with late Holocene hydroclimate variations at Owens Lake. The Olancha Dunes are the primary dune area in the southern sector of the Owens Lake basin (Lancaster et al. 2015) and consist of approximately 4.39 km² of sparsely vegetated low dunes with isolated ridges that extend nearly 5 km southwest from the historical shoreline of Owens Lake. The Olancha Dunes are composed of poorly to moderately sorted fine to medium-sized sand (Figure 4-4B; Lancaster

et al. 2015). Sand from the southern portion of the Owens Lake basin is mineralogically mature, with the Olancha Dunes having the highest quartz content compared to other dune areas (Figure 4-4A; Lancaster et al. 2015). As at the Keeler Dunes (discussed above), the high enrichment of quartz in the Olancha Dunes area can be attributed to either abrasion of feldspars during aeolian transport or weathering of feldspars during transport or phases of dune stabilization (Muhs 2004) followed by immersion and transport in lacustrine environments during historic water-level fluctuations in Owens Lake. Sand in the southern part of the Owens Valley basin, especially near the Olancha Dunes, was likely sourced from the South Sand Sheet (Figure 4-10) during prehistoric low stands of the Owens Lake. The South Sand Sheet is hypothesized to be sourced from a system of ephemeral washes draining the Coso Range, including Vermillion Canyon (Lancaster et al. 2015). Bacon et al. (2020) dated two paleo-shorelines at elevations of 3,638 ft (1,109 m) and 3,635 ft (1,108 m) and obtained luminescence ages of 11.5–45.6 ka. To date, we are unaware of other constraining ages. Using augers to complete an OSL coring campaign across the sand dunes/sand sheets could also provide a comprehensive set of young ages that are reflective of recent dune activation. While quartz would be the primary target of these studies (as the dominant mineral and best suited to produce higher-resolution OSL ages), feldspar may also provide helpful OSL ages and distinctive luminescence characteristics that allow for increased confidence in the OSL chronology.

The resultant aeolian drift potential derived from recent decades of wind data near Olancha (Lancaster et al. 2015) indicates a strong bimodal sand transport regime at the southern end of the Owens Lake basin with opposing northerly (29 percent frequency) and southerly (51 percent frequency) modes. In recent decades, net resultant drift at Olancha is toward the north/northwest (345 degrees), although a secondary reversed transport mode toward the south/southwest also occurs. Similarly, the aeolian sand drift regime at the nearby Dirty Socks site indicates bimodal transport with a northerly (2 degrees) net resultant sand drift potential vector, but with a more balanced frequency of opposing northerly and southerly transport modes (42 percent for each). As a result, the present connection between sediment supply from the Owens Lake bed, particularly from the South Sand Sheet area (Lancaster et al. 2015) to lake-marginal dunes at Olancha, is not as clear as it is on the northeastern side of the lake at Keeler. Although the resultant sand drift potential vector and crescentic dune slip faces in the Olancha Dunes indicate net northward sand transport, it is also apparent in historical aerial imagery (Figure 4-20) that the dune field has expanded toward the south, particularly since the 1970s. Between 1977 and 1986, unvegetated areas of the southeastern portion of the dune field expanded approximately 0.3 miles south toward Fall Road and a private residence evident in the imagery (see area enclosed by yellow line in 1986 image in Figure 4-20). It is possible that this trend partly reflects the secondary south/southwest mode in the sand transport regime resulting from increased sediment supply or availability. This sediment supply or availability could have resulted from the drained lakebed or other anthropogenic activities at the Olancha Dunes, but this cannot be distinguished with existing monitoring data.

The 1970s and 1980s correspond to an era of increasing off-highway vehicle (OHV) activity within dune fields in California both generally and specifically in large recreational preserves, such as the Oceano Dunes State Vehicular Recreation Area and the Imperial Sand Dunes Recreation Area. Research at these sites shows clear associations between OHV activity, decreasing vegetation cover, and increased dust emissions (Cheung et al. 2021; Gillies et al. 2022; Groom et al. 2007; Walker et al. 2023). In the Owens Valley, the Bureau of Land Management (BLM) manages an OHV recreation and primitive camping area of approximately 400 acres (1.6 km²) or 36 percent of the total area of the Olancha Dunes, which has been operational for decades. Although there is little to no research on the trends in vegetation cover within the Olancha Dunes or on the associated impacts of decades of OHV activity on dune development and dynamics, it is probable that the extent and duration of OHV activity at the Olancha Dunes could explain recent declines in vegetation cover and dune expansion, particularly in the southern area and other areas of active camping and riding.

NORTH OF THE LAKE

As shown in Chapter 3, there have been a few exceedances observed at Lone Pine or other northerly monitors like North Beach and Lizard Tail between 2017 and 2024. The District’s exceedance database suggests that many of the off-lake exceedances at these monitors are from anthropogenic disturbances (e.g., landfill) or from regional sources

outside the OVPA. However, some evidence shows that certain regional events were augmented by sources north of the lake. These include a regional exceedance on 05/11/2018 that was “augmented by local sources north of North Beach between the monitor and the Lone Pine landfill,” a regional exceedance on 10/27/2019 that was “augmented by local sources north of North Beach monitor,” and a regional exceedance on 10/11/2021 that was affected by “additional sources as the front traveled down the valley.” Additionally, as demonstrated in Chapter 3, sources near the Fort Independence monitoring site may be primary causes of a few high PM₁₀ concentrations observed there.⁴ However, as discussed in Chapter 3, it is unclear if these sources are derived from active disturbances of the surface (e.g., construction, highway traffic), or if they are from other off-lake sources of dust that are the focus of this report. Given that the panel was tasked to consider “expected future changes in off-lake dust sources over time,” potential future sources of dust emission to the north of the lake are explored here, even though it is not clear that these sources are currently resulting in exceedances. These potential sources include small depressions found inset in the land surface, vegetated areas found throughout the valley floor including agricultural fields (active and abandoned), as well as areas of former alkali meadow between Lone Pine and the north edge of the OVPA.

Microplayas

A considerable number of small areas, mapped as “microplayas” by the District, exist in the area around North Beach within 3 km of the historical shoreline (Figure 4-21). These mapped areas likely do not all share the same origin; they may be depositional, erosional, or a combination of the two. For example, some could be paleo-lake deposits that have been exhumed by wind (Stone et al. 2000). Alternatively, these landforms could be wind erosion hollows that subsequently filled with heavy-texture material from local overland flow (Stone et al. 2000). Other examples of these landforms sit at the toe of fans and are likely purely depositional features. At present, there is no evidence to show that microplayas are anthropogenic in origin, but they could be classified as such if human activity led to the wind erosion (e.g., through vegetation disturbance or groundwater disturbance that impacted vegetation) that created these local erosional depressions. Nonetheless, amid otherwise relatively coarse soils, these landforms do have more potential to contribute to PM₁₀ exceedances compared to other areas of the uplands, especially because they appear to have lower vegetation cover compared to the surrounding areas. Nonetheless, there are currently no exceedances obviously associated with these surfaces, and it is unlikely that they will be major source of exceedances in the future.

Vegetated Areas North of the Lake, Including Areas of Former Alkali Meadows

As described in Chapter 2, large alkali meadows historically existed throughout the Owens Valley (Benson et al. 2002), but much of this groundwater-dependent vegetation was lost, and at present, the dominant vegetation type in the Owens Valley is characterized as arid or semiarid scrub (LADWP and County of Inyo 1990a). When loss of groundwater-dependent vegetation like this occurs, meadow vegetation is frequently replaced by desert shrublands dominated by xeric shrubs such as saltbush (Atriplex spp.), sagebrush (Artemesia tridentata), and blackbrush (Coleogyne ramisissima). The result is a conversion of a meadow with greater than 30 percent cover (often considerably more) to scrublands with 10–20 percent cover (Figure 2-5; Manning 1997). Vest et al. (2013) showed nearly an order of magnitude increase in aeolian transport between alkali meadow and scrub sites, and the fluvial and lacustrine sediments across the OVPA valley floor (Bacon et al. 2006) provide ample saltation and suspension-sized material. Thus, the panel investigated the potential for historical alkali meadows to produce PM₁₀, even though based on the analysis of data in Chapter 3, it is not certain that this is currently resulting in exceedances at existing monitoring sites.

As described in Chapter 2, historical anthropogenic activities such as the increase in groundwater pumping, and surface water in-valley use and export to Los Angeles after 1970 impacted vegetation along the OVPA valley floor. Between 1986 and 1992 there was a considerable but variable decrease in groundwater levels in the Owens Valley

___________________

⁴ The U.S. Environmental Protection Agency (EPA) does not use this monitor to assess attainment with the PM₁₀ National Ambient Air Quality Standards (NAAQS).

as a result of this use (Vest et al. 2013), which was associated with significant vegetation change at those locations that later required revegetation projects to reduce dust emissions (LADWP 2023). For example, LADWP (LADWP 2023 pp. 3-14 to 3-15) states, “Fluctuations in water tables due to groundwater pumping have caused approximately 655 acres of groundwater-dependent vegetation to die off [at Hines Spring South]. …The goal will be to restore as full a native vegetation cover as is feasible, but at a minimum, vegetation cover sufficient to avoid blowing dust will be achieved in that area.” Further evidence of historic emissions induced by groundwater withdrawals was found in the highly wind-eroded Rindge soils that were visited by the panel during the May 29–30, 2024, meeting south of Independence. Evidence of decimeters of historic erosion by wind was evident in a location that was at the site of a former groundwater seep, as evidenced by the high organic matter content and clear histic epipedon. Thus, the panel finds that areas north of the lake within the OVPA, including historical alkali meadows, have produced dust in the past and thus have the potential to produce dust in the future.

Irrigated agriculture removes native vegetation and changes the soil structure, which can lead to dust emission, especially if these agricultural lands become abandoned (Birmili et al. 2008; Field et al. 2010). As described in Chapter 2, there was a reduction in the amount of irrigated acreage of Los Angeles–owned land from the mid-1960s to 1970 (LADWP and County of Inyo 1990b). Abandoned agricultural land does not have the same cover of native perennial vegetation as similar areas that were not plowed, and thus they generally produce greater amounts of

dust (Zucca et al. 2022). A photograph from 1950 (Figure 4-22A)⁵ shows extensive dust emission from the area north of Owens Lake in the vicinity of Independence, likely from disturbed land or active, fallowed, or abandoned agricultural fields.

A groundwater management plan was implemented in the 1990s (see Chapter 2) along with revegetation projects on abandoned agricultural land and land that was impacted by the loss of groundwater-dependent vegetation. LADWP (2023) notes that many of the revegetation projects are “implemented and ongoing,” whereas others are “complete,” and still others are “fully implemented but not meeting goals.” However, the panel did not attempt to determine if the completed projects have sustained their target vegetation cover nor whether these levels are sufficient to suppress dust emission.

As described in Chapter 2, the panel expects considerable future change in climate in the region of the Owens Valley. Warming will drive a large loss of snowpack, a dramatic shift in runoff time to earlier in the wet season, and an increasing “flashiness” of runoff (Alex Hall, personal communication, July 2024; Harpold et al. 2015). Measured and projected increases in temperature (Abatzoglou et al. 2021; Williams et al. 2019) have and will continue to increase evaporative demand, affecting all types of vegetation. Increasing evapotranspiration will drive reductions in surface runoff, even in scenarios with increasing precipitation (Owens Valley Groundwater Authority 2021). Future changes in precipitation are predicted to bring longer-term drought conditions with intermittent extreme wet events (Swain et al. 2016; 2018). Extended drought could impact areas of natural non-groundwater-dependent vegetation, areas of former groundwater-dependent vegetation, or former irrigated agricultural fields, because rainfed vegetation remains susceptible to drought. Projecting into the future, a combination of climate changes (e.g., decreased runoff, increased evapotranspiration, increased drought) could result in increased dust emission from vegetated areas north of the lake, especially if there are changes in groundwater management agreements or land management policies. Revegetation projects that have not yet met their revegetation goals may be especially susceptible to future changes in climate because such areas may have soil/climate conditions that do not favor rapid revegetation; in such places, a climate less amenable to vegetation growth is likely to slow progress even further.

___________________

⁵ See https://www.soaringmuseum.org/hof_more.php?id=117.

CONCLUSIONS AND RECOMMENDATIONS

The panel considered multiple lines of evidence to infer the origins and evolution of major PM₁₀ sources in the OVPA or those that might become important sources in the future. One important process the panel considered was “winnowing.” This process suggested that PM₁₀ material from the dry lakebed that may have been deposited onto off-lake landforms would be expected to decrease over time due to its resuspension and progressive removal by aeolian processes. This hypothesis was based on a correlation between on- and off-lake exceedances at the Dirty Socks monitor between 1999 and 2012, which has not been borne out with more recent data for estimated emissions trends and exceedances. Most current PM₁₀ emissions from off-lake areas are likely not a result of resuspension of PM₁₀ material that was deposited on off-lake landforms from the lakebed. Instead, the presence and common replenishment of highly emissive flood deposits provides ample fine particulates that can be emitted as PM₁₀ as long as the horizontal flux of saltation-sized particles is sufficient to emit dust from the surface.

Northeast Side of the Lake

The northeastern side of Owens Lake is host to several landforms including the Keeler Dunes and the Slate Canyon/Keeler Alluvial Fan Complex that have overgone major changes over the last century and are substantial contributors to exceedances. During the 20th century, the Keeler Dunes transitioned from a largely vegetated dune system that was stabilized by greasewood (Sarcobatus vermiculatus) to an active dune field. The emergence of Keeler Dunes as an active dune field resulted in abundant saltation that can drive PM₁₀ emissions from flood deposits that are continually replenished from the alluvial fan. Therefore, ongoing PM₁₀ exceedances from this area are a direct result of the destabilization of the Keeler Dunes. The panel finds that the net transport direction, available imagery, and the evidence for groundwater-dependent vegetation currently present in the dunes supports the conclusion that increased sand transport following the diversion of water from the Owens Lake destabilized the Keeler Dunes. Changes to surface hydrology resulting from construction of berms above Highway 136 appear to have had an impact on upland (non-groundwater-dependent) vegetation but are unlikely to have led to the destabilization of groundwater-dependent vegetation in the Keeler Dunes.

Several constructed berms northeast of Highway 136 were intentionally designed to alter surface hydrology, directing overland flow to specific points of discharge points along the highways. The panel did not analyze each berm-related flood deposit, but instead considered the berms on the Keeler/Slate Canyon Fan as potentially representative of similar features around Owens Lake. These berms have had appreciable, localized impacts on the distribution of flood deposits in the Keeler Dunes region, especially following impacts from the remnants of Hurricane Kay in September of 2022. Further investigation would be needed to determine the impacts that these berms have on the potential for off-lake PM₁₀ emissions from flood deposits, if any.

Southern Side of the Lake

The southern side of Owens Lake is host to the Olancha Dunes and multiple alluvial systems that are important sources of PM₁₀ exceedances. The scientific literature on the origin and evolution of the Olancha Dunes is quite sparse, but what available evidence does exist indicates that the dunes formed prior to the diversion of water from Owens Lake. The panel’s analysis shows that the dunes experienced a slight southward extension (approximately 0.3 miles) from 1944 to current day. This southward extension could be the result of increased sediment supply following the diversion of water from Owens Lake or from other natural or anthropogenic activities. Olancha Dunes is also the location of an OHV and dispersed camping recreational area that makes up approximately 36 percent of the total dune area. There is little to no research on the impacts of decades of OHV activity on PM₁₀ emissions at Olancha Dunes, but research from other sites like the Oceano Dunes State Vehicular Recreation Area and the Imperial Sand Dunes Recreation Area shows clear associations between OHV activity, decreasing vegetation cover, and increased dust emissions. Additional study using aerial photography, Portable In-Situ Wind Erosion Lab dust emission potential measurements, and the Bureau of Land Management’s records of impacts from recreational activity could provide information on the contribution of recreational activity to PM₁₀ emissions.

The southern side of Owens Lake also hosts multiple alluvial channel/wash systems that deliver and rework sediments from the neighboring Coso and Sierra Nevada ranges. These alluvial channel/wash systems supply sand and PM₁₀ material, and they can only support low-density vegetation cover, which create conditions ripe for high PM₁₀ emission. While the replenishment of these alluvial systems is a natural process that has been occurring for millennia, anthropogenic alteration of the flowpaths through the constructed infrastructure may change the amount and distribution of impounded water and sediment and therefore change its potential to contribute to PM₁₀ emissions. Climate change is projected to make extreme precipitation events more frequent and intense, which would more frequently replenish fine sediments in flood deposits that contribute to PM₁₀ emissions.

POTENTIAL SOURCES NORTH OF THE LAKE

Current data indicate a stable shallow groundwater table in the area around Owens Lake. However, there is substantial evidence that areas to the north of the lake have seen decreases in vegetation cover, which may

have contributed to historical dust emission in the OVPA. A groundwater management plan and a number of revegetation projects were implemented in the 1990s to reduce blowing dust in affected areas. If this land is not managed carefully, dust emission from the area to the north of the lake could increase, especially under changing climate conditions.

FURTHER RESEARCH TO ESTABLISH THE ORIGIN AND EVOLUTION OF OVPA SOURCES

More chronological research may reduce uncertainties surrounding the origin and evolution of dune fields and flood deposits. Collecting sediment samples across dunes and flood deposits by coring or auguring may be the best way to collect a comprehensive set of data. These methods may illuminate relatively recent processes that occurred after the diversion of water from Owens Lake, construction of berms, and the rerouting of Highway 190.