As of 6:00 pm today, the water level in Suicide Basin was still rising. According to our estimates, current water volumes in Suicide Basin are in the ballpark of the 2016 flood. Today at noon, the median of our Monte Carlo simulations yielded 0.0325 km3, while the volume of the 2016 flood was approximately 0.035 km3.

Without a gage at Suicide Basin, we currently estimate the water level elevation by extrapolating the water level increase from the most recent warm period (1.08 m per 24 hours). According to this extrapolation, we reached 437 m at 3:00 pm today, which was most likely an overestimation.

The real water level was probably around 436 m, which is supported by photographs taken by a NorthStar pilot earlier today (Fig. 1). Overestimation occurred because of the relatively low temperatures and minimal precipitation since last week’s warm period.

The lowest point of the glacier dam will control when and at what elevation the lake will spill over (unless the lake will drain via subglacial conduit). This point is found along the glacier boundary (Fig. 1). As our current drone-derived DEMs do not extend all the way to this point (Fig. 2), its elevation has to be estimated. I did this by determining the low point on a 2016 satellite-derived DEM, followed by a correction for glacier thinning. This yields an elevation between ~440 and 442 m. An extended drone survey is planned for our next site visit.


At the current rate of water level increase, the lake in Suicide Basin will start spilling over in three to six days. If the water level increases faster, it may happen within the next two days. If the lake spills over, the water will start flowing along the glacier boundary. The water may then start incising the ice, which in turn will increase the water flow and thus the rate of ice incision (positive feedback). We would be surprised to see the water flowing along the glacier boundary over a very long distance. Instead, we would expect it to find its way into the glacier at one of the depressions visible in digital elevation models (which are typically crevasses or moulin-type holes). Predicting the timeline of this process is challenging, especially because we have not observed it at Suicide Basin previously. However, we would suspect this process to unfold more slowly than a drainage event via subglacial conduit.

Note that although the lake is close to spilling over, draining via subglacial conduit may still occur instead (a combination of processes may occur as well). For this scenario, we would expect a few days between the onset of drainage and peak runoff at Mendenhall River, as observed in previous years. Peak levels at Mendenhall Lake reach nearly 12 feet in our preliminary flood prediction runs.