May 31, 2011

Papers I'm reading: Trends in the timing and magnitude of floods in Canada

GIS map of the Atlas of Canada, showing numerous hydrometric measurement stations in the western half of the country
In case anyone hasn't noticed, Canada is huge. The country's vast area and varied topography lends itself to a multitude of different hydrological regimes, influenced atmospherically on a meso-macro scale by Polar & Tropical air masses. The warm vs cold air mass war has its battleground across the latitudes of Canada, from 42° - 66.6°N, with the only region that can be called "calm" being the ice-capped tundra islands north of the Arctic Circle. Very few places in Canada do not receive decent snowfall, so spring freshet runoff from snowpack melt is typical, and measuring it is every Canadian fluvial hydrologists nitty-gritty.
Bella Coola airport during Sept. 2010 flood
A prime example of a rainfall-induced flood

Climate change towards a warming trend must have an impact on the various hydrological regimes and how the hydrologic cycle has been altered due to that impact. As I've said, there are a multitude of different hydrological regimes, so there are likely a multitude of different trends. This line of thought brought me to an academic paper in the Journal of Hydrology that reviews trends in timing & magnitude of floods in Canada due to hydrologic shifts, and does so by looking at the established physiographic regions of Canada. Juraj Cunderlik and Taha Ouarda from the Natural Sciences and Engineering Research Council of Canada Chair on Statistical Hydrology analyzed flood data gathered from several dozen strategically placed monitoring stations throughout the hydrometric network of Canada.
This study investigates trends in the timing and magnitude of seasonal maximum flood events across Canada. A new methodology for analyzing trends in the timing of flood events is developed that takes into account the directional character and multi-modality of flood occurrences. The methodology transforms the directional series of flood occurrences into new series by defining a new location of the origin. A test of flood seasonality (multi-modality) is then applied to identify dominant flood seasons. Floods from the dominant seasons are analyzed separately by a seasonal trend analysis. The Mann–Kendall test in conjunction with the method of pre-whitening is used in the trend analysis. Over 160 streamflow records from one common observation period are analyzed in watersheds with relatively pristine and stable land-use conditions. The results show weak signals of climate variability and/or change present in the timing of floods in Canada during the last three decades. Most of the significant trends in the timing of spring snowmelt floods are negative trends (earlier flood occurrence) found in the southern part of Canada. There are no significant trends identified in the timing of fall rainfall floods. However, the significance of the fall, rainfall-dominated flood season has been increasing in several analyzed watersheds. This may indicate increasing intensity of rainfall events during the recent years. Trends in the magnitude of floods are more pronounced than the trends in the timing of floods. Almost one fifth of all the analyzed stations show significant trends in the magnitude of snowmelt floods. Most of the significant trends are negative trends, suggesting decreasing magnitudes of snowmelt floods in Canada over the last three decades. Significant negative trends are found particularly in southern Ontario, northern Saskatchewan, Alberta and British Columbia. There are no significant trends in the magnitude of rainfall floods found in the analyzed streamflow records. The results support the outcomes of previous streamflow trend studies conducted in Canada.
What did they find? (delving into the abstract's details)

The study first defines some ground rules:
  • Minimum of 20 years of observational data
  • Statistically significant flood regimes usually have bimodal seasons (spring freshet season + fall rainfall season). There were no stations identified with three or more significant flood seasons.
  • Watersheds used in the statistical study are characterized by relatively pristine and stable land-use conditions, with less than 5% of the watershed area being modified by human development
  • It is not feasible to get thorough coverage of streamflow data across Canada due to the country's sheer breadth, and inaccessibility of certain regions
  • The pulse day = the day for which the cumulative departure of the streamflow from the average streamflow for the year is most negative. Most acutely observed during the spring freshet season within high relief terrain
    I'm not going to detail their establishment of a statistical methodology to gauge multi-modal flooding seasons, as I'm more interested in the findings about floods in the country (statistical know-how is something I'm lacking, soon to be remedied). In terms of timing trends, readings in southern Canada over the last generation are showing a trend towards earlier spring melt floods. This trend is particularly acute in southern Ontario, Alberta, and BC. Changes in freshwater ice break/freeze up spring dates are strongly linked to large-scale atmospheric and oceanic oscillations, with positive feedback mechanisms as the driving force. Only 16 stations, which is 10% of those included in the study, had a statistically significant trend in the timing of spring freshet floods during the last generation. Of those 16, 14 had a negative trend pointing to earlier occurrence. Only 2 stations had a positive trend pointing to later occurrence of spring floods, and those were in the tundra of Nunavut.
    The real crux of the study is the findings about the change in flood magnitudes, as the values associated with volume/discharge are more striking than the shifts in the timing of said discharge. There is only a weak trend in having earlier melt runoffs, but a more significant trend in having a lower magnitude of those melt runoffs. The trend shows decreasing streamflow in heavily glaciated areas (ie. alpine glaciers of Rockies, valley glaciers of sub-Arctic & western Cordillera) during the typical spring runoff season for those areas of April - July. On average, the mean annual spring maximum flows have decreased by approximately 1% per year over the analyzed period of 1974 - 2003. My home province of BC had some intriguing findings: there was some significant increasing trends in spring maximum flows (increased snowmelt-induced flows) and significant decreasing trends at the beginning of summer (reduced snowmelt-induced flows); this is essentially showing that the snowpack is melting at a greater pace earlier in the season, leaving the summer months less supplied with meltwater from the snowpack. This is not a good trend as that scenario exacerbates drought conditions typical of BC in the summer, when a persistent Pacific High settles over the region. Indeed the paper finds that accompanying climate data (air temps) are shifting at a quicker-than-anticipated pace from the coldest to the warmest season, causing greater flow spikes in response.
    You mentioned "bimodal". What about the fall rainfall season trends?
    Almost a non-issue amongst the observed data. The authors note only a potential for increasing rainfall during the fall season, as certain regions recorded occasional above-average spikes in fall rainfall in the last decade. Further accumulation of data throughout the coming years will shed light on the significance of the rainfall floods, and trends in their timing & magnitude. I should note that not all physiographic regions in Canada that have bimodality have the second, lesser, rainfall-induced one in the autumn months. Rather, some regions, most notably the prairie provinces, have the rainfall-induced flooding in the summer months due to convective storms.

    What are the take-home messages?

    There are many implications for the decreasing snowpack melt, and how it interacts with certain biogeochemical cycles. Particular regions of Canada rely on the spring melt to supply freshwater, and in BC, hydroelectric power will be adversely affected by a decrease in discharge at certain expected intervals. You've likely seen the news on how floods in Manitoba have adversely impacted populations living in flood plains there, but that has more to do with latitude, topography and a South - North sloping drainage basin (Anne Jefferson at Highly Allochthonous has a great post on the topic). Ongoing changes in the timing and magnitude of spring floods are not restricted to a particular flood seasonality type, but rather occur across Canada in different climatic and hydrologic regimes. Ultimately, the paper highlights that trends in the magnitude of floods are more pronounced than trends in the timing of floods, and the changes in magnitude are having a great impact among natural systems connected to the melting of the snowpack in the spring.

    Additional Info:

    May 23, 2011

    Accretionary Wedge #34: That is Weird

    This months Accretionary Wedge is being hosted by Dana Hunter over at En Tequila Es Verdad, which Google translates as "In Tequila is Truth" (don't ask this teetotaler what that means). The theme is any geology which the blogger considers weird. My limited experience means many geological phenomena I observe are initially head-scratchers, but subsequent investigation usually becomes a learning experience, and later I can't imagine a time before understanding the phenomenon.

    In choosing what to post about that's weird, a memory popped into my head that quickly settled the issue. What better to include in this carnival on weird geology than a landform that has no unifying theory on its origin, but rather a bunch of hypotheses? What I speak of is the Mima Mounds, located 20km south of Olympia, Washington. This set of mounds is the only I have visited, but variations of Mima Mounds exist elsewhere: Lake District in Oregon, Northern China, and in the Western Sahara, to name a few. The geographic and climatic spread means that certain groups of the mounds have more explanation for their genesis, ex. the Oregon-based mounds have a more definitive volcanic morphology. But the mounds in Washington I'm focusing on continue to baffle geologists. Hypotheses about their origin range from animal construction - seismicity - periglacial kettle topography.

    The mounds outside Olympia measure around 5-8 feet in height, and 12-20 feet in rough diameter. They are similar to the prairie-based pimple mounds of the southern Midwest states, and their might be connections based on pedological similarities. If you are a keen geologist, each explanation will stir up good probing questions, many of which are yet to be answered fully. For instance, the gopher proposal is criticized for lack of zoological evidence at the Washington Mima Mounds, plus their density raises questions of competition for food resources if a multitude of gophers built them, or necessity if many mounds were built by a few gophers. The earthquake hypothesis is a compelling one, and more research into the physics of the mounds' granular material should be revealing once it comes forth.
    Click for larger version and read about the different hypotheses scientists have for the origin of the mounds.
    Info board courtesy Washington State Department of Natural Resources
    Most current research into the Mima Mound phenomena is concentrated around the periglacial hypothesis: Kettle & Kame topographical depressions (sun cups) were filled with glacial sediment during rapid retreat at the end of the last Ice Age. Repeated outburst floods as the glacial front retreated & disintegrated provided sediment that filled the depressions. Those depressions experienced subsequent freeze-thaw heave; hence they are arguably akin to small-scale pingo formations. However, glacial conditions are not apparent at several places where the mounds exist, even when examining deep-time paleogeography. Also of note is that not all mound formations have the same soil/sediment profile, even within the same mound group. Some of my own quick observations at the Mima Mounds Natural Area include how the mounds are more diversely vegetated, some mounds have a deflated appearance, and that exposures of the substrate revealed a primarily gravelly/pebbly mixture that reminded me of glacial diamicton.
    Washington DNR LIDAR image of Mima Mounds (left) with matching Google Earth image (right)
    Site is near Littlerock, Wa. (46° 53.273'N 123° 3.054'W)
    Geology that can be considered 'weird' is refreshing to have around. A lot of it is nature's abstract art. I considered doing the tessellated pavement structure of Eaglehawk Neck in Tasmania, but that has a thorough explanation, and honestly, when thinking of strange geological formations, one's that are unexplained and/or under debate strike my fancy more. Finding out that not everything in earth science is yet definitive gives me a chance, albeit small, to be a future pioneer.
    Additional Info:

    May 17, 2011

    Field photo Set #3

    A recent stopover in Princeton, BC gave me an opportunity to sidetrack to a couple interesting outcrops that expose sedimentary strata of different formations. It was a perfect day - clear skies, temps in the high teens, dry but not too dry - summer come early. Without even trying, on occasion I was within meters of adventurous local wildlife - a beaver, several deer, and a young black bear.

    One particular exposure along the TCT, nicknamed the "Red Ochre Bluffs" because they were used by natives to create red pigment, is particularly interesting for its very reddish color, due to a high % content of the mineral haematite (iron oxide) within the bedded chert. Technically it is an outcrop of the Vermillion Bluffs shale member, part of the Allenby formation of the Eocene epoch (45-50 Ma). This is a fossiliferous member, where fossils of maple, alder, fir, pine, dawn redwood and ginko have been found, along with one of the world's oldest fossilized bees. There is a noticeable dip to the beds of about 10°, striking NNE-SSW.

    At another part of Princeton, behind a small restaurant, is the only exposure in Princeton of the Summer Creek sandstone member of the Allenby formation:
    A keen eye will notice the concretions, the cross laminations, and an apparent conglomerate boulder 'xenolith' that became part of the package, though it might be a beaten up granite-family rock. The sandstone layers have an approximate dip of 25°, and a strike of E-W. The member is mantled by a foot of glacial till (Princeton has a few kettle lakes of interest that showcase glacial geomorphology). This sandstone is some of the toughest I've felt; the layers are highly compacted and the presence of plenty of cementing material makes it a strong variant.

    The Vermillion Bluffs exposure can be found @  49° 26.695'N 120° 32.665'W, after a 2km walk along the TCT. Part of the walk goes through a long tunnel where you can practice your bet megalomaniacal laugh.
    The Summer Creek sandstone rockface can be found @ 49° 27.313'N 120° 30.646'W, behind Billy's Family restaurant, where you can park and take a look at how First Nations hollowed out a cave within the sandstone to store plunder.

    Additional Info:

    May 13, 2011

    Looking through the archives: Melbourne's climate

    Satellite image of pyrocumulus clouds over Victoria & New South Wales,
    taken by NASA's Aqua satellite in early February 2009
    The climate of Melbourne, Victoria, Australia is one of great variation relative to the rest of the continent, but still moderate as oceanic effects and broad relief limit extremes, especially in terms of freezing temperatures and excessive rainfall. The city is located at the south-central side of the state of Victoria, which is the southeast corner of Australia. With a Latitude/Longitude of 37.5°S 145.0°E, Melbourne falls within what is technically considered the mid-latitude region. In terms of global circulation patterns, the city is affected by not only mid-latitude westerlies, which brings air across the west of the continent to the city, but also by the subtropical High and the Antarctic circumpolar vortex. The subtropical High normally resides in New South Wales for most of a typical year, but during the southern hemisphere summer the ITCZ (Intertropical Convergence Zone) descends into northern Australia, thus pushing the subtropical High south into Victoria and Melbourne. These seasonally changing patterns make Melbourne a diverse region in terms of weather, and fosters extreme events such as bushfires & droughts that are amplified by El Niño southern oscillations.

    Melbourne at a Glance

    Melbourne lies quite flat on the horizon. A coastal city of 4 million with a secluded port as its access to the Indian Ocean (via Bass Strait), Melbourne is at the confluence of two major rivers that flow into Port Philip (Yarra and Maribyrnong). Geologically, Melbourne is mostly underlain by Silurian marine sediments, and modern alluvium from Yarra. The marine sediments were uplifted from the shallow Bass Strait. This highlights how low the general relief of southern Victoria is. With a sea level decrease of just 70m, a land bridge would form between the city and the island of Tasmania.

    In terms of precipitation, this low relief makes Melbourne susceptible to flash flooding during more intense showers/thunderstorms in both La Niña and spring seasons. Poor drainage and infiltration through city streets, combined with a low greenspace ratio, exacerbate flash floods in the city.

    Click to

    May 9, 2011

    USGS and GIS

    I can't stand acronym overload, but lately I've been caught overusing them. Yesterday a discussion about construction at a park and the info they put up about ephemeral streams resulted in me using 'WRC', 'CWH' and 'BFR' to the befuddlement of my friend. But in the case of this blog, I can't imagine anyone reading it not grasping the two acronyms in this posts title.

    Many of the earth science/physical geography professors and grad students I've talked to are almost unanimous in enjoying the field work aspect of their research. However, some of them are apprehensive about GIS technology, believing that increased ubiquity of the software will reduce or even eliminate the need for hard data gathering that is a big part of field work. Is that true? I don't know. I can surmise that depending on circumstances it could be, and I've certainly been to a few geomatics presentations where the crux of a new technique is to gather empirical geographic data without leaving the desk.

    However, good GIS data is hard to find, and free GIS data even harder. Our own Geological Survey does not release much data to the public for free, and finding any GIS digital elevation model (DEM) files is akin to a needle-in-haystack search. So I was surprised when looking up information on Crater Lake and stumbling across USGS's collection of DEM files on the lake's bathymetry. Scratch that ... I wasn't surprised, as I've gotten used to the ridiculous restrictions my own government places on what should be freely available public information (these restrictions aren't limited to the geographic realm). One of my GIS bosses remarked on how his students are finding it really hard to collect digital data for their term projects, and he further mentioned that most projects involve analysis of geographic phenomena in BC. Even my own work projects involving examinations of local watersheds was shelved due to lack of shapefiles or anything that could be converted to such without a lifetime of eye damage.

    But is the USGS stuff I found really any good? Cracking open ArcGIS to take a look at the bathymetry data, I was immediately struck with déjà vu. Where had I seen this layout of Crater Lake before? Of course, I see a variation of it in passing everyday, pinned up on my apartment wall:
    At a gift shop in Crater Lake park, there was a huge 6-foot tall version of the above geologic map with even more detail. Alas, it was only for display, but the 4-foot tall version available for customers has good detail for a pretty penny.
    I was able to do lots with the data, modifying it, adding annotations, creating a color scheme for entities, and segmenting portions of Crater Lake geology using primitives as splitters. Using the data in ArcGlobe was especially interesting, as you get a true sense of scale and perspective for all the features above & below lake level because you can strip away the water with a click. Some of the diagrams I've seen in Crater Lake academic papers have definitely used these DEM files. The resolution is incredible, as the 2000 bathymetric survey conducted by USGS, NPS, and University of New Hampshire's CCOM used multibeam acoustic sounders that translated data to a 2-meter-per-pixel representation factor.
    The virgin view of Crater Lake in ArcGlobe [Top Left]. USGS 7.5 arcminute DEM of Crater Lake, showing shaded relief bathymetry [Top Right]. Angled perspective of 7.5 arcminute Crater Lake DEM, with annotated structures [Bottom]
    Thank goodness for the USGS! Good physical geography GIS data is out there, based on surveys they've conducted, and data they digitized. Best of all, it's freely available to the public. Now if only my federal government would allow GSC to follow suit.

    On the subaqueous features/structures of Crater Lake

    Rhyodacite Dome = youngest (5 Ka), shallowest subaqueous feature. It is a highly silicic lava dome that has formed from a vent that intruded through both the andesitic Central Platform & the eastern flank of Wizard Island. Rhyodacitic flows and domes were common in Mazama's history between 40-5 Ka.

    Central Platform = Subaerial andesite flows that experienced magma differentiation from within the primary chamber. Successive eruptions (7-6 Ka) that built up the platform were above lake level, giving the platform a low, broad relief until flows met the old shoreline.

    Merriam Cone = Cinder cone whose lower portion/foundation was formed as a resurgent dome, composed of subaqueous andesite, formed from completely submarine eruptions circa 7.7-7.5 Ka. It has a nearly geometrically idealized cone shape. Peak is ~150m below current lake level. Origins and formation still not agreed upon (Resurgent dome or Cinder cone or how much of a combination??).

    Phantom Ship = Oldest (400 Ka), partially subaqueous feature. Considered to be the top of a basic-andesitic volcanic dike, a remnant of a small vent that might have fed parasitic cones on pre-cataclysmic Mazama's eastern flank (see also Devil's Backbone).

    Various Depositional Basins = Colluvium, volcaniclastics, volcanic breccia, and ash particles, all weathered and transported to the low-lying depressions of the caldera. The rim has plenty of talus slopes. There is a huge landslide deposit on the southern end of the lake, called the Chaski Bay landslide, with a debris avalanche volume of 0.1 km3 (couldn't find an exact date on the slide). Evidence of the slide triggering a mild tsunami is found at Cleetwood Cove on the northern end.

    Additional Info:

      May 3, 2011

      Columbia Basin Trip: Day 3 - McCall and a bunch of Falls

      My third and final day out in Oregon included a spectacular hike through a nature preserve, and a drive along an historic highway where there is the greatest concentration of high waterfalls in the US. I lost count of how many waterfalls I saw, and recalling all their names is tough (Elowah, Oneonta, Multnomah, Horsetail, Ponytail, Latourell, Sheppard Dell, just to name a few). Each fall had its own character, in which the basalt rockfaces they flowed over varied in layering, formation, and vegetation.
      Info board @ Rowena Crest parking

      But first, before the falls, it was time for a satisfying hike. Hiking Oregon's Geology guide book spoke of a 4-5 mile hike through a nature preserve just outside The Dalles (which I learned is pronounced "Dals", not "Dall-ehz"). Called the "Rowena Crest/Tom McCall Preserve", most of the geology relating to this hike is distant, observed by viewing the Columbia River gorge on the opposing Washington state side. Their side has the Ortley Pinnacles, fragments of basalt cemented together at a fault line. Mounds dotting the preserve's open stretches of grassland are erosional remnants of St. Helens ash deposits; they are more subdued and more sporadic than the infamous Mima Mounds, which may or may not have the same genesis. The real highlight of the hike is ecological, as at this point in springtime I could spot lupines, balsamroot, and blossom trees, and that was with my weaker-than-novice botanical background. The whole scene was very colorful at this time of year, and was capped off with sunny, turbulent weather.

      The bulk of the geology in Day 3 came in the form of neverending pullovers and short, grinding switchbacks to view a multitude of waterfalls. This collection of falls was created when the Missoula floods cut away the gentle foothills of the flood basalts, leaving the cliffs and their creeks-turned-falls. If my recollection is correct, all but two of the falls I visited were of the plunge variety (Oneonta was Step-Pool, Sheppard Dell was tiered/fan). Discharge was excellent, fueled by mid-spring rains and freshet melt from peaks ranging from 1200-1600m elevation (see above cross-section).

      Two particular falls were a real treat to behold: Multnomah & Latourell. The former is the highest in Oregon, second highest in the continental US, and technically is a hybrid of plunge/step-pool, with a higher and lower set of falls. Differential cooling rates of the various Columbia flood basalt lava flows (Grande Ronde Basalt) provided Multnomah with distinctive layering from top-bottom. I could make out entablature basalt layers (fast cooling, fractured into irregular blocks & joints), a pillow basalt layer (fastest cooling when exposed to water, forming rounded cobbles), and columnar basalt layers (slower cooling under entablature, forming slender hexagonal blocks) at the very top and interspersed near the bottom behind the rockfall scarp. Multnomah Falls splashwater erodes softer layers of rock below & behind the falls, creating a plunge poll and amphitheater semi-cave. The higher falls recede upstream faster than the lower falls due to weaknesses in their lowest basalt layers. Large pieces of basalt rockface have historically been calved off, including a 400 ton piece falling into the plunge pool and drenching a wedding party 15 years ago.
      "... Observations of waterfalls over Columbia River basalt have shown that falls often occur where flows are flat lying or dipping upstream. This condition allows blocks produced by vertical joints to remain stable until support is withdrawn by erosion of softer interflow material at the base of individual flows. The rate of erosion of interflow areas probably largely controls the rate of retreat of the falls. The amphitheater-shaped valleys common to many of the falls within the Gorge are due to the freeze-thaw action of water from the splash mist that has penetrated the joints. ..." [Norman and Roloff, 2004]
      Latourell Falls from the lower gallery
      Click for short video of falls in action

      Latourell Falls were nearly as impressive as Multnomah, and benefited from not having the trappings of tourism that's part of the Multnomah stop. Though not as high and not as layered, Latourell was pristine & photogenic, with lichen giving an entablature formation a splash of color, and the columns on the bottom undercut layer looking like an arrangement of cathedral organ pipes. The 76m plunge is the most unfettered of all high waterfalls in the gorge region, as others tend to impact (horsetail) at least slightly against the vertical rockface. It's one of those beauties where it's hard to take a bad picture.

      I'm not really good with coda's, but I can say that there is one dominant feeling I came away with from this trip = I want to go again...but somewhere new. And as gas prices continue their northward march, the western US states are looking even more preferable than they were beforehand. I certainly got to experience a huge slice of basalt geology, which was quite the contrast from the granite geology of my home base. I think the next time calls for a true desert locale, something not really available in my home province (Osoyoos doesn't count). Ahh the possibilities...I just wish they wouldn't butt up against the lack of time & money.

      Additional Info:

      May 1, 2011

      Columbia Basin Trip: Day 2 - Sisters along the Columbia

      Looping around the Wallula Gap early the next morning, I made a stop at an interesting formation that at first makes me think 'volcanic plug'. Alas, the Twin Sisters are not so, but rather they are the erosion-resistant last vestiges of great flood basalts that covered this particular part of the Columbia Basin. The pillars, which are shaped like irregular molar teeth, consist of vertical remnants of pillow basalt atop a foundation of columnar basalt. Certainly there are other basalt forms in the Gap that also have bits of these remnants still standing, but the Sisters are the largest, most impressive, and least covered in vegetation. As you can see above, their interesting shape spurred native legends & provided some flavor to the geology.
      Two distinct forms of basalt make up the Sisters

      Every geoscience geek with the smallest bit of knowledge of Pacific Northwest geology knows of the Glacial Lake Missoula floods and their offspring, the Channeled Scablands. The Wallula Gap, its numerous flood basalt flows, and the Twin Sisters themselves were all shaped by the late Pleistocene floods that ultimately found their outlet down the Columbia River. Akin to the Umtanum anticline mentioned on Day 1, compression of the basalt layers created an anticlinal ridge in the Wallula Gap area. An ancient river, precursor to the modern Columbia, slowly but surely cut a gorge through the layers as the gradient increased. Thus we had another water gap created thanks to the steady pace of uplift matching erosion. Of course, once the Missoula outburst floods began, tremendous volumes of water swept down towards the Gap where they were constricted, and thus erosive power was mostly focused on widening the Gap.

      For those not familiar with the Scablands, their features & their origins: During the late Pleistocene, the Cordilleran ice sheet advanced into northern Idaho, Montana, and Washington. Gigantic ice dams were formed behind lobes of the continental glacier, holding back thousands of cubic kilometers of meltwater. When these dams broke, huge amounts of water were unleashed, following the path of least resistance through eastern & central Washington, constricting at the Wallula Gap, then funneling down the Columbia River, creating the Columbia River gorge. The last of the floods, called the Bretz flood, released 1600 km3 of water in a two-day period, inundating nearly the entire Channeled Scablands region, and even extended into the Willamette River valley south of Portland, before exiting into the Pacific near Astoria.
      Diagram of geographical interaction between Pleistocene ice sheets (blue), Glacial Lake Missoula (yellow), and the full extent of the Channeled Scablands (orange)

      I lingered at the Twin Sisters for quite some time on the quiet weekday morning, and I wandered around looking at various perspectives of the palisade basalt, those exposures of basalt along ridges that make it look as if the area is fortified. Ahead of me was a long drive westward on the Columbia River interstate highway, made longer by tough crosswinds picking on my little Yaris. I didn't get to witness or scrutinize much more in terms of geology on Day 2, but the picturesque drive was superlative, and I did notice an interesting phenomenon about the Columbia river that tweaked my hydrologic bone...
      Looking west on the Columbia River along I84, just outside of Rufus. Those are whitecaps, not rapids
      Winds were gusting up to 70 kph, making it hard to open the car door, but I had to snap a photo of the waves on the Columbia going against the current. That's right, against the current, upstream, eastward. This isn't abnormal or against the laws of physics. All I can see as an observer is the surface of the river, and out of the two forces acting on the surface, the winds are winning...on the surface. Who knows how far down the water depth column the winning force becomes the downstream current? 3, 4, 5 feet, out of hundreds of feet? I couldn't exactly whip out a dingy and a current meter and head out into the frenzied waters, though I wish I could. In any case, interesting food for thought as I travel towards the multitude of high waterfalls that would be the 3rd day of my Columbia Basin trip.

      Additional Info: