Prehistoric Hanging Canals of the Safford Basin

Prehistoric Hanging Canals of the Safford Basin

Loading Social Plug-ins...
Language: English
Save to myLibrary Post page as Slide Download PDF
Go to Page # Page of 7

Description: Recent studies have investigated the scale and scope of a remarkable series of late classic prehistoric (est 1350 AD) canals literally "hung" on the edges of steep sided bajada mesas. Taken in stone age context, these seem to represent incomparably superb and exceptional world class engineering. .

Author: Don Lancaster (Fellow) | Visits: 2721 | Page Views: 2749
Domain:  High Tech Category: Business Subcategory: Engineering History 
Upload Date:
Short URL:


px *        px *

* Default width and height in pixels. Change it to your required dimensions.

Maxims of Tech: Rules of Engagement for a Fast Changing Environment
Prehistoric Hanging Canals of the Safford Basin
Don Lancaster Synergetics, Box 809, Thatcher, AZ 85552 copyright c2010 pub 06/10 as GuruGram #108 (928) 428-4073

ome recent studies have investigated the scale and scope of a remarkable series of late classic prehistoric ( est. 1350 AD ) water management structures found in Arizona's upper Gila Valley. These canal systems are characterized by their being literally "hung" on the edges of steep sided, gently sloping mesas formed from remnant Quaternary age bajadas. The mesas today will appear characteristically rocky, sparsely vegetated, thinly soiled, and largely infertile. At places, the hanging structures are placed as much as 60 meters above their surrounding drainage basins. It is quite clear that the highest feasible points on mesas were carefully selected for canal routes "on purpose". In several cases, shorter and easier direct drainage routes seem to have been clearly avoided. Five or more distinct hanging canal systems have been identified. Most of which trend from northwest to southeast. Eventually leading northward to the Gila River valley proper. System lengths are estimated to total twenty kilometers or more. These systems seem to nearly totally exploit a number of adjacent Mount Graham fed perennial streams that include Jacobson, Marijilda, Rincon, Deadman, and Frey canyons. Whose present stream flow rates are typically in the fractional to low CFS ranges with seasonally higher peaks. The systems appear clearly distinct and separate from other extensive but more conventional prehistoric area canals based upon fertile bottomlands and Gila River and tributary water sourcing. Associated with the systems are above-grade aquaducts of significant height and length. Along with apparent elaborate methods of purposeful switching of the water routes between major delivery drainages. In sum, these systems appear to represent a major understanding and a careful exploitation of hydraulic fundamentals. Typical canal cross sections might be one meter wide by twelve centimeters deep.
-- 108 . 1 --


Their use seems to be initially associated with mesa top agave cultivation. Perhaps followed by more conventional downstream water reuse. Assisted somewhat by historic maintenance and rebuilds, several portions of the canals still flow to this day. Other portions of most of the systems remain largely pristine. Although currently filled with fine grained and loess-like or sedimentary depositional sands. Present ownership is typically Arizona State Lands and currently remains largely undeveloped or exploited. Although of often rather difficult access, major canal portions are usually easily traced and clearly identifiable. There are few access roads and fewer linking trails. Especially on the mesa tops. The canals often create the illusion of "water flowing uphill" in that the chosen mesa top slope is usually somewhat steeper than the rate of fall of the canal system itself...

After reaching a mesa top through a long, gentle, and an apparently carefully calculated optimal grade, and then continuing as far as seems possible along the characteristically flat but gently sloped mesa top, the canal systems will typically "fall off" the far end of their mesa in steep but apparently highly controlled and nondestructive cascades. Some eventually ending up in areas characterized by habitation sites and ag structures such as grids, mulch rings, field houses, linear features, or trincheras.
-- 108 . 2 --

Most systems seem to have a "breakaway" initial diversion point. At which major flood damage would appear to be easily and quickly repaired without significantly impacting the main structural portions of the canals themselves.

Some Examples
From South to North, the systems can be arbitrarily named Ledford, Marijilda, Rincon, Deadman, and Robinson. These might be examined in more depth...

The Ledford system starts at the John's Dam and climbs "up" onto the remnant bajada mesa just north of Metate Peak. It then passes Goat Tank, Ledford Tank, and similar structures still apparently in use for modern stock raising. Several of the switching arrangements can be noted that provide steep cascading descents on south to various valley floor tanks. There does appear to be considerable modern use of this resource. The final and easternmost "dropoff" point is quite obvious owing to rather extensive parasitic vegetation. The Ledford canal system appears to be the least well explored and interpreted of the hanging canals and remains difficult of access. Some speculation exists as to whether prehistoric hydraulic features were also exploited further to the south, based on Spring Canyon, Veech Canyon, and in Stockton Wash itself. While the terrain seems clearly advantageous, no obvious features are known to have been observed. Except for a few short sections of deep vee ditches, larger modern diversion channels, and canals that seem to be historic or modern. The area remains of considerable interest.

Marijilda Canyon
Earlier and upcanyon portions of the Marijilda system have long been rebuilt by historic pioneers. These concrete lined ditches now deliver water from reliable sources in Marijilda Canyon to seasonal Lebanon irrigation storage reservoirs. The rebuilds are comparable in size and depth to a modern final delivery cotton field ditch. And of a classic deep vee shape. Which is quite distinct from the prehistoric originals elsewhere. The rebuilt system route deviates from one branch of the prehistoric just at the watershed crossover between the Marijilda and Rincon drainages. At that point, the eastern reach goes over a significantly above grade aqueduct. One that is roughly 1.5 meters high and as much as 100 meters long. It does not seem immediately clear what is gained by the aqueduct, compared to a slightly lower surface route.
-- 108 . 3 --

The aqueduct exactly follows a watershed ridge and consists of solid fill. There is no apparent provision or need for cross flow under that canal at this point. Its use and need remains enigmatic. The eastern reach then goes into its hanging canal mode, sloping "up" onto its mesa and reaching the top a kilometer or so later. This portion is dry and largely full of fine dirt particles. Largely making up what today is a nearly ideal hiking trail. Cross sections can be evaluated at numerous points due to some erosion cut throughs. These usually suggest a fractional meter width and a ten centimeter depth to the water channel. Occasional brush directly along the canal route suggests a minimum of fifty or more years of disuse. In many places, the mesa slope forms the upcanyon wall, while a constructed rock barrier takes care of the downcanyon side. Construction consistently shows an extreme economy of energy, with the fewest possible rocks and the least amount of dirt moved in a minimralist manner. There seems to be no obvious evidence of use of draft animals or iron scrapers or wheelbarrows, or anything of that nature. Nonetheless, countless thousands of man hours over long time intervals appear to have been involved in engineering and building these highly unique systems. Presumably, water was hand carried or otherwise diverted to mesa top needs. After running as far as practical along the mesa top, the water system steeply drops off a cascade system. And eventually returns to modern storage resovoirs in the valley floor. Numerous valley branches then continue to various ag sites further downstream.

Rincon Canyon Area
The Rincon Canyon area is an extension upon the Marijilda system. At its northern extreme, it climbs back "up" onto a mesa and in two large sweeping loops, then doubles back upon itself. Making a "U" turn and then heading up canyon in a reverse direction. The canal here is somewhat wider than normal, with its upmesa wall set by the gently sloping bajada, and a single rock wall usually retaining water on the downmesa end. An earlier habitation site is nearby, and numerous ag structures are present just off canyon in the form of grids, linear features, field houses, and mulch rings. The canal in this region is quite distinct in its web based satellite photography. Eventually the canal goes once again into a shorter and lower mesa edge hanging mode, this time returning north to the valley floor while heading off the bajada.
-- 108 . 4 --

Deadman Mesa
The initial portion of the Deadman Mesa hanging canal has only been postulated to date, having been seemingly replaced by a modern pipeline. Nonetheless, the route follows the characteristic pattern of being hung on a steep wall. In this case, a canyon wall is used rather than a mesa edge. After a one kilometer "climb", the optimally graded system reaches a modern pond. At that point, historic water was diverted down a side canyon through transite pipes to become part of the presently unused Frey Mesa water system. Just beyond obviously modern headgates, vee weirs, and concrete structures, the canal continues down the Deadman Mesa remnant bajada. The presumed prehistoric continuance is absolutely devoid of modern materials and techniques. And consists primarily of two rows of guide rocks half a meter to a meter apart. The water route is typically along the highest feasible portion of the mesa. This canal reach remains fully functional and often flows to this day. Following a two kilometer run, the canal traverses an extremely narrow portion of Deadman Mesa. There is an apparent three way water switch at this point whose appearance is rather obvious on satellite imagry of the area. Owing to parasitic vegetation along the water routes. Switching water routes seems to consist simply of moving rocks around. No headgates or modern diversion structures are involved. The northernmost switching routes water down a steep canyon into Porter Springs tank. From there, the water flow apparently continued to known and rather dense prehistoric habitation and ag sites. These sites are also characterized by a large number of small peripheral check dams. Some of which have downstream aprons, and others of which a full double width splash containment. Sizes appear consistent with plant nurseries. Significantly, prehistoric occupation seems centered upon the probable water route. Locations to the north and south are largely and conspicuously absent of prehistoric development. There are fascinating but unverified and unproven hints of canal extensions that go beyond the habitation areas. Other ag features in the area include grids, linear features, and mulch rings. The center Deadman switching once routed water to Lower Deadman Tank and is largely unused and dry today. No obvious evidence of prehistoric occupation has been found or verified in this immediate area to date. The southern Deadman switching currently routes water to Upper Deadman Tank. Significant prehistoric cultural resources are present in the Rincon Canyon area below this tank.
-- 108 . 5 --

Robinson Ditch
This ag structure was named for a historic Mormon pioneer who apparently restored, maintained, and improved the canal system. Arguments that the structure was in fact initially prehistoric include (A) Its astonishing similarities to known prehistory in adjacent canyons; (B) Mesa top bajada routing more consistent with prehistoric needs and goals; (C) A total absence of concrete, iron, headgates, or more modern techniques; and (D) The size, depth, rock relocations, and energy levels required appear totally consistent with stone age technology. The Robinson Ditch starts off with an apparent three-way routing switch in Frey Canyon. At this point, water can be routed "up" the ditch onto a bajada remnant mesa, down past Sheep Tank into Spring Canyon, or follow its traditional route down Frey Canyon. After "rising" to the mesa top and continuing along its narrower portions, the canal rather steeply descends to end up very near the Spring Canyon route after a unique run of nearly five kilometers. Modern Frey Dam overflow usually routes through Spring Canyon. Most of the Robinson Ditch is presently full of loess or water borne sediments and in need of restoration. Large parasitic trees adjacent to the hanging canal portion appear to be long dead...

-- 108 . 6 --

Current work
As with most Southwestern Archaeology, any funding is sorely limited. As is the manpower needed for further study and interpretation. Dr. James Neely, professor emeritus at the University of Texas at Austin, is a long term researcher here. One of his many earlier papers appears here, and another here. Some further publications are in process. Studies are ongoing. Additional champions and support are urgently needed. A crucial present issue is an accurate mapping of the entire area to acceptable resolutions far better than what is readily web available. It is possible that one or more Draganfly Drones might be suitable for this task.
A web published and open sourced detailed master index of all prehistoric ag features in the area would also seem to be highly useful.

There are serious problems, outrageous costs, poor maintenance, and sharply limited availability involving the existing directory systems. The web has fully guaranteed that their attempts at extreme secrecy simply no longer work and are clearly no longer applicable. It is also only a matter of time for web based general aerial photography to make the needed 10:1 further resolution increase required for nothing to remain either hidden or hideable. A key question remains "Why their manic obsession with the bajada?" Also needing addressing are successful methods of precisely dating and isolating differences between prehistoric, early pioneer, CCC ( Civil Conservation Core ) and modern constructs. Perhaps CSI forensic techniques may emerge of use here. For instance, "would CCC fingerprints survive on the undersides of rocks?" Resolution of that question should be a sure fire winner for a Master's Thesis in any of a dozen of fields. Other obvious questions involve the instruments used to guarantee optimal canal slopes. As pure speculation, perhaps some "floating boat" scheme might have been used. Or some variation of a right angle sight being added to a plum bob system. No known examples appear to have survived. Today, the Gila Valley is well noted for several examples of outstanding high technology. Most obviously involving telescopes and new significantly "greener" energy efficient mining techniques.
When taken within the context of available stone age tools, techniques, and energetics, these prehistoric grids and hanging canals clearly illustrate many examples of comparably superb and exceptionally world class engineering.

-- 108 . 7 --