AGU 2018 presentation

Integration and Analysis of Lidar and UAS Surveys

to Study Coastal Sand Dunes Evolution and
Changes in Field Microtopography

Helena Mitasova
A. Petrasova, V. Petras, J. Jeziorska, C. Kling

Link to this presentation:https://ncsu-geoforall-lab.github.io/dynamic-surfaces-talk/agu2018.html#/

Data acknowledgment: NOAA digital coast, NC Floodplain mapping, NCALM, USACE JALBTCX, USGA/NASA/NOAA coastal mapping, NC NGAT UAS mapping

Multitemporal elevation data

Lidar surveys along North Carolina coast since 1996
UAS and SfM made 3D mapping widely accessible
monitoring dynamic landscapes is now possible at increasingly finer spatial and temporal resolutions
rapid evolution of technology leads to data with varied properties: often we have to work with multitemporal data rather than time-series

Lidar-based DSM: 1996 and 2008, differences in coverage, accuracy, point density

Hardin, E., Mitasova, H., Tateosian, L., Overton, M., 2014, GIS-based Analysis of Coastal Lidar Time-Series, Springer Briefs in Computer Science, Springer, New York, 84 p.

GRASS GIS temporal framework

Efficient management and analysis of multitemporal 2D/3D raster and vector data

Gebbert, S., Pebesma, E., 2014. A temporal GIS for field based environmental modeling. Environmental Modelling and Software 53, 1-12.

Jockey's Ridge sand dunes

Largest active dunes on the East coast, state park with static boundaries
Management challenges: wind blown sand covers structures and roads outside the park, dune lost elevation

Mitas, O., Mitasova, H.; Brothers, G., Weaver, K., 2014, Managing Dune Landscape Changes at Jockey's Ridge State Park, North Carolina, Tourism in Marine Environments 9(3-4), pp. 155-167.

Coastal multitemporal elevation data

photogrammetry: 1974, 1995, 1998
lidar: 1999,2001,2007,2008,2009,2012,2014,2015 (7/2018,10/2018)
structure from motion from UAS imagery: 2016,2017

1974 1998 1999

2001 2009 2015
vertical / horiz. accuracy: 0.7 / 0.4 m photogrammetry, 0.15 / 0.2 - 0.8 m lidar

Processing DEMs

1m resolution DEMs: interpolated with regularized spline with tension
Elevations along I-158 were extracted to evaluate systematic errors
UAS survey did not cover the road, GCPs on dune were used

number of surveys per grid cell profiles along I-158 road section from lidar DEMs 1999-2015

Processing DEMs

1m resolution DEMs: interpolated with regularized spline with tension
Elevations along I-158 were extracted to evaluate systematic errors
UAS survey did not cover the road, GCPs on dune were used

number of surveys per grid cell profiles along I-158 section: corrected lidar DEMs 1999-2015

DEM time series visualization

Jockey's Ridge 1974 - 2017: southward migration at 3-6 m/yr, landform transformation from crescentic dune to sand starved, fast moving parabolic dunes

DEM time series visualization 1974 - 2017

Post-hurricane Mathews ripples

Orthphoto from UAS survey, october 2016

Annual dynamics from Planet imagery

Satellite imagery at 3m resolution, September 2017 - July 2018, captures impact of storm in March 2018 with large ripples similar to those observed post Mathews

Planet: world’s largest constellation of Earth-imaging (micro) satellites providing daily observations for entire Earth at 3m resolution

Complex pattern of changes

DEM differences show both loss and increase in elevation
total sand volume is stable (~3.5 mil cubic m)
volume of core (sand not moved between 1974-2017) is only 30% of total volume

Overlayed DEMs 1974 - 2017: crossections

Landform change: contours time series

Dune gradually split into smaller parabolic dunes: contours capture the landform change but they are hard to read

16m

20m

Space-Time cube visualization

DEM time series is converted into space-time voxel model in TGRASS and evolution of a contour is represented as isosurface: 16m and 20m

Peak elevation change

linear trend in loss of peak elevation at 0.3 m/yr, from 43 m to 20 m
accelerating horizontal migration

Location of peak migrates following dune transformation

Peak elevation change

growth between 1917 - 1949, reconstruct DEM from 1949 and 60s imagery
accelerating horizontal migration

Location of peak migrates following dune transformation

Jockey's Ridge story

The 43 m high dune was a transient landform,
transitioned between forest and active dunes over the past 2000 years

Dune in early 1900 and in 2016, sand removal at the park boundary in 2003

Agricultural field monitoring

UAS based monitoring of field microtopography and crop topography
28 UAS surveys, 2 lidar surveys
4 different UAS platforms (Trimble UX5, eBee Plus, DJI Inspire, Phantom 3)
Agisoft Photoscan, GRASS GIS temporal framework
fusion with lidar for watershed modeling
used in courses and for student projects

Jeziorska, J., Montgomery, K., Mitasova, H. 2018, Microtopography and Crop Vigor Changes Assessment Using Time Series of UAS Derived Data. AGU Fall 2018 meeting poster B33F-2378
Petrasova, A., Mitasova, H., Petras, V., Jeziorska, J., 2017, Fusion of high-resolution DEMs for water flow modeling, Open Geospatial Data, Software and Standards 2(6)
Jeziorska, J; Mitasova, H; Petrasova, A; Petras, V; Divakaran, D; Zajkowski, T., 2016, Overland flow analysis using time series of sUAS-derived elevation models, ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., III-8, 159-166. DOI:10.5194/isprs-annals-III-8-159-2016