Skip to main content

News & Events

Grad student talk - Using high-temporal-resolution, repeat terrestrial LiDAR to compare topographic

Thursday, November 07, 2013, 4:30PM - 5:30PM


Theo Barnhart


RL-1 room 269

Full title

Using high-temporal-resolution, repeat terrestrial LiDAR to compare topographic change detection methods and to elucidate the hydrometeorologic controls on the retreat rate and form of the Selawik Retrogressive Thaw Slump, northwest Alaska


Retrogressive thaw slumps (RTS), a type of catastrophic thermokarst indicative of permafrost degradation, are forecast to increase in frequency and magnitude with a warming climate. RTS flux a disproportionate amount of sediment and nutrients to downstream ecosystems with the potential for adverse impacts. Characterizing the processes and hydrometeorologic drivers through which these features grow is necessary to better understand how these features may behave in the future. The Selawik RTS initiated in 2004 and has grown at a rate of 7-20 m/yr. In 2012, the feature was 200 m wide with a vertical headwall ~20 m high at its apex. We use a 58 scan repeat, ground based LiDAR data set collected over the summers of 2011 and 2012 and interval camera imagery to:

  1. compare two topographic change detection methods, cloud to mesh (C2M) and the Multiscale Model to Model Cloud Comparison (M3C2) algorithm.
  2. compare the error analysis techniques used with C2M and M3C2, (3) describe RTS mass loss processes, and (4) investigate the drivers of RTS retreat rate and form.

We found that C2M reports higher magnitude topographic change over short time periods (~12 hours) and lower magnitude topographic change over long time periods (~20 days) when compared to M3C2. The spatially variable error analysis protocol used with M3C2 better accounts for the sources of uncertainty in point cloud data sets used for topographic change detection than C2M. TLS data from 2011 show a diel pattern in the mean retreat rate of the feature while data from 2012 show a more mixed signal. These differences are likely due to the warm, dry conditions experienced in 2011 verses the cool, wet conditions experienced in 2012. Statistical modeling indicates that RTS retreat rate and form are most sensitive to net radiation (R2 27.4%, pVal: 0.001 and R2 82.0%, pVal: <0.001, respectively). We interpret this to indicate that spallation-type mass loss processes, driven by elevated net radiation, are most effective at removing material from interstitial ice dominated RTS features. Furthermore, we find that hydrologic pathways in the tundra upslope of the feature control the cuspate form the Selawik RTS headwall.