The resource link to the left links to Part 1 (Introduction) of 7 of this presentation. Further parts are linked below.
Dr. Chris Hopkinson introduces a series of seven videos that covers the topic of LiDAR and its application in water resources planning and monitoring. In the first video, the technologies and the fundamental scales and mathematics of LiDAR are described.
In the second of seven videos, Dr. Chris Hopkinson describes the data that comes back from a LiDAR survey. He points out that the point cloud attributes are different as the radar pulses come back from different forms for LiDAR scanners. He explains multiple returns, primary returns and secondary returns and how that supplies more information in the point cloud data. Canopy penetration is important when ground elevation modelling is objective. Landscape surfaces are one model product (canopy height modelling in forestry, building modelling in urban design). Digital elevation modelling requires last returns but last returns are not necessarily ground returns. He finishes by discussing return intensity and how it can be used to interpret surfaces and moisture contents. Colour imaging and LiDAR can be combined for additional surface interpretation.
In the third of seven videos, Dr. Chris Hopkinson describes the uncertainties that come as LiDAR point cloud data is measured, selected and then used in data modelling. Considering the number of steps in processing LiDAR data, data uncertainties needs to be considered and accounted for in planning the LiDAR data acquisition. High data density is not necessarily a good thing since the trade-off is lower radar intensity. Signal intensities do vary and decimetres of error can be introduced if not properly calibrated. He finished this video by discussing how in forestry, canopy modelling needs some years of data before you can be confident that the LiDAR is measuring real changes in the forest vegetation.
In the fourth of seven videos, Dr. Chris Hopkinson describes the typical workflow that is followed in processing the raw returns data into a point cloud and raster modelling to typically produce a digital elevation model. Tiling of data is required to batch process portions of the total project. Aircraft position needs to be accounted for. Quality control for flight line overlaps needs to be considered. Ground control for the project needs to be reliable and properly interpreted. Classifying the point cloud is next. DEM and canopy height models and leaf are index are output modeled products. Rasterization usually requires interpolation of LiDAR data, usually by aggregation. Dense forests and the search radius can create data voids. The raster method does impact modelling results.
Part 5 - Hydrological applications
In the fifth video of seven, Dr. Chris Hopkinson describes the opportunities of LiDAR in hydrological modelling. LiDAR currently may not be used for all these applications but the opportunity is present. GIS based flow routing from LiDAR does run into problems when roads and other transportation "barriers" are part of the watershed. Urban drainage networks also need extra information in addition to the LiDAR DEM. Ephemeral stream crossings can be identified from LiDAR; not so easy in the field. Contouring approaches misses topography under trees so LiDAR provides better watershed estimations. He finishes this with a short description of Alberta's wet areas mapping initiative.
Part 6 - Hydrological modelling
In the sixth video of seven, Dr. Chris Hopkinson discusses the uses of LiDAR in hydrological modelling. Optical data and LiDAR data can be combined for detailed vegetation classification. Wind and solar aspects can be interpreted from LiDAR (e.g. aerodynamic roughness). Cross sections information is a place where LiDAR can significantly improve hydraulic modelling. Although surveyed cross sections provide precise elevation data the advantage of LiDAR is to fill in cross sections where regular surveying would be expensive. Likewise in multi-channels, LiDAR slope data would help in defining multiple channel flow. Flood inundation scenarios are mentioned at the end of the video.
Part 7 - Monitoring applications
In the final video of seven, Dr. Chris Hopkinson discusses the use of LiDAR for water resource monitoring. He in particular talks about snow pack monitoring in mountain watersheds. In the alpine zone, he found good correlation between LiDAR differences and snow pack depths. LiDAR confirms maximum snow pack volume is at treeline. The details from the LiDAR study allows translation of these results to other adjoining watersheds. LiDAR also shows the runoff contribution of ice cores buried in moraines after glacier recession. LiDAR in remote areas provides useful information about hydraulic gradients on complex river systems (the example is the McKenzie River delta). He shows how LiDAR picks up hydraulic jumps and allows the modelling of detailed river hydraulics that otherwise would require detailed stream gauging.
Dr. Chris Hopkinson's presentation was part of the LiDAR/SAR workshop at the University of Lethbridge, Alberta, Canada, June 26-27, 2014 organized by the Alberta Terrestrial Imaging Centre at the U. of Lethbridge. Dr. Hopkinson is the Research Chair for the Centre. The Alberta Land-use Knowledge Network was able to record many of the presentations.