Knowledge of the process and time scales for the scour, burial, re-exposure, migration, and subsequent re-burial of unexploded ordnance (UXO) that remain in and on the seafloor in coastal environments are key to the planning for the future use of these sites, including possible remediation. Because these processes are intrinsically chaotic, predictions of the fate and transport of a broad-field UXO population are extremely sensitive to the initial state of that population, specifically the numbers of UXO per cubic meter of seabed sediments, the distribution in the depth of impact burial, and the initial horizontal distribution of the UXO population.

There were two primary technical objectives for this project: (1) integrate a ballistics impact model after Chu and Ray (2006), Hale (2009), and Chu et. al. (2010) with the presently configured VORTEX UXO Mobility Model (MM), and subsequently validate with UXO survey data; and (2) use the integrated MM (referred to as UXO MM) to answer two charge questions posed by the ESTCP review panel at the Spring 2013 In-Progress Review Meeting:

1. For projectiles deposited in Lake Erie in a variety of initial conditions (fully buried on impact, partially buried, on the surface), how many years does it take for the effects of the initial conditions to be washed out?
2. Do the position and orientation of the projectiles only become randomized during severe storms, or are the normal seasonal weather patterns sufficient?

The UXO MM is a processes-based model that uses VORTEX computational methods to generate three-dimensional (3-D) simulations of subsequent burial, exposure, and migration of complex UXO shapes. In these simulations, the model accounts for effects of large-scale erosion or accretion of the seabed (far-field processes) and fine-scale vortex shedding, scour, and bedform evolution around the UXO shape (nearfield processes). Farfield processes are those that alter the seabed elevation over length scales that are comparatively large with respect to the size of an individual UXO round. Nearfield processes are due to the flow disturbance caused by the UXO and affect the seabed elevation by local scour as well as induce hydrodynamic forces that cause the UXO to move. The present MM software was validated in two ESTCP-funded field tests (MR-201003 and MR-200417), where it correctly predicted all the basic behaviors of UXO test surrogates with high quantitative predictive skill factors.

The ballistic impact burial model that was embedded into the MM architecture is STRIKE35, a six-degree of freedom (6-DOF) ballistics model developed by Chu, et. al. in 2010. STRIKE35 is a derivative of the Navy’s well proven mine impact burial model IMPACT35. STRIKE35 is written in MatLab and was transposed to FORTRAN and integrated in the MM flow chart as a separate module as part of the integrated UXO MM. STRIKE35 has physics for the three basic processes of impact burial mechanics: (1) aerodynamic trajectory through the atmosphere after Hume (2007) and Hale (2009); (2) impact with the air-water interface after Chu and Ray (2006); (3) free-fall through the water column after Chu and Ray (2006); and (4) impact with a sedimentary seabed after Chu and Fan (2007). The computational sequence proceeds round by round and the model output includes both depth of impact burial and orientation of the round on impact with the seabed. STRIKE35 was validated during controlled impact studies in two test ponds at the Naval Air Warfare Center, Weapons Division (NAWC/WD), Indian Wells, California. STRIKE35 is an evolution from the mine impact burial model IMPACT35 that was validated during the Office of Naval Research’s (ONR) Mine Burial Program. The MBES (predecessor expert system to the Underwater Munitions Expert System [UnMES]) was validated during ONR’s Mine Burial Program.

The UXO MM was used to predict migration and burial behavior of UXO rounds in Lake Erie impact areas off Camp Perry, Ohio. From the Camp Perry firing logs, the likely range activity gives an estimate that 3.9 million UXO targets have impacted the offshore regions of Impact Areas 1-3. Direct comparisons of UXO “targets” found by underwater magnetometer surveys conducted by SAIC versus UXO MM simulated impacts were good, with a coefficient of determination for range distance to impact of R-squared = 0.868 out to distances of 12 km from the firing pits, and a skill factor, Rh, calculated at Rh = 0.86 for impact burial. In the main range fan, 92% of the targets checked by the SAIC divers were actually UXO, while the remaining 8% were other metal objects. The size of the UXO varies from small caliber bullets to as large as 106 mm rounds. There also was at least one 250 lb bomb found near West Sister Island, but by far the dominant UXO types are the Army 106 mm M344 round artillery shells, the 60mm M49A2, and 81 mm M43A1 mortars. A probability distribution of the impact burial depth of these 1.3 million impacts simulated from the firing logs is clearly bi-model, with the majority being deep impacts on the order of 60 cm to 120 cm resulting from high apex trajectories with javelin entry modes. The UXO rounds associated with these deep impacts are likely not subject to subsequent mobility due to the fetch limited wave heights and wave periods along the southwest shores of Lake Erie. However, the smaller peak in the bi-modal distribution is attributable to shallow impacts associated with oblique trajectories and oblique skip entries. These shallow impacts (burial depths ~ 15 cm) amount to 2.53% of the firings in the Camp Perry firing logs, and account for about 98,613 of the UXO targets in Impact Areas 1-3. Due to these shallow burial depths, 98.6% of targets from oblique skip entries become exposed (97,000 exposed rounds), and these UXO become the focus of the UXO mobility analysis.

During 28-year UXO MM simulations, only 4,485 UXO rounds become exposed more than 50% and subsequently have the potential to migrate from their point of impact due to natural causes; nearly all of those rounds are in a near-shore area between the beach berm and the -3.35 m Low Water Datum (LWD) depth contour (extending offshore about 700 m away from the beach berm). While the net movement of these mobile UXO is small, ranging from tens of centimeters to several meters, the gross movement can be large, as much as several thousand meters. If the shore zone off Camp Perry were to remain stable, it might take as long as 25 years for these mobile UXO to reach a final steady state distribution. However, the shore zone is not stable and has historically been erosional with progressive shoreline retreat. It is this long-term erosional trend that threatens to expose UXO along the shoreline for the indefinite future, where human interaction may become problematic.

The UXO MM is extremely data intensive with its input and gridding requirements, and requires relatively long run times for any given UXO scenario involving a particular munitions type, initial burial state, bottom type, and environmental forcing. Regardless, the successful implementation of this model at Camp Perry, Ohio, revealed several important issues that may require further use of this model at this site in the event of another extreme Lake Erie storm. Presently, about one-quarter of the 97,000 rounds predicted to become exposed in 28 years by the UXO MM simulation are due to shoreline retreat. The fundamental distinction, however, is that shoreline retreat is progressive, not episodic like extreme storm events or lake water level variation, and not periodic like seasonal beach profile variation. Consequently, an average of about 980 rounds per year could be exposed by shoreline retreat, and these are new rounds, not previously subject to visual detection or beach clean-up. That number will continue to grow over time until shoreline recession exhumes all of the deepest nearshore and beach dune impact burials off the FUDS Beach at Camp Perry. This finding indicates that future UXO surveys and monitoring efforts should focus on the erosional nearshore zone, beginning on the barrier dune where land-based detection assets can be used, and extending offshore to the -4 m LWD depth contour where high resolution electromagnetic and acoustic detection systems can be used. Offshore of this erosional nearshore region, the UXO field is stable and the majority of UXO will remain permanently entombed based on their initial ballistic impacts.