1 00:00:00,030 --> 00:00:05,220 um but what I've been looking at is how bedrock rivers are transporting coarse 2 00:00:05,220 --> 00:00:10,349 sediment and I'm looking at a series of rivers across Taiwan one of the things 3 00:00:10,349 --> 00:00:14,309 that you can see in these drone videos is that the rivers are incredibly 4 00:00:14,309 --> 00:00:18,359 variable so we have these incredibly coarse sediment this is up to 10 meters 5 00:00:18,359 --> 00:00:23,070 in diameter um and actually I'm gonna come on scale on screen in a moment 6 00:00:23,070 --> 00:00:30,150 first scale oh there I am and so you can see that there's spatial patterns in the 7 00:00:30,150 --> 00:00:34,489 grain size we also have patterns and where it is relative to the banks and 8 00:00:34,489 --> 00:00:40,050 patterns relative to bedrock exposure as we move down these different rivers so 9 00:00:40,050 --> 00:00:43,350 one of the things that we're interested in is how these bedrock rivers are 10 00:00:43,350 --> 00:00:47,489 transporting sediment and are roading the banks and eroding into the bedrock 11 00:00:47,489 --> 00:00:51,809 and we need to look across the full range of variability and understand how 12 00:00:51,809 --> 00:00:57,030 is this changing over space and also how is this changing over time so what are 13 00:00:57,030 --> 00:01:02,420 the variability in the spatial locations of sediment transport capacity and also 14 00:01:02,420 --> 00:01:06,620 how is this changing with different flows different flow events 15 00:01:09,640 --> 00:01:10,880 so the way 16 00:01:10,890 --> 00:01:14,480 that we're approaching this is with a series of drone surveys so on the left 17 00:01:14,490 --> 00:01:18,720 is the drone survey from the field and on the right is the photo realistic 3D 18 00:01:18,720 --> 00:01:23,430 model that we can actually make of the exact same site so these are not quite 19 00:01:23,430 --> 00:01:27,330 exactly aligned but you can see that these are highly photorealistic they're 20 00:01:27,330 --> 00:01:31,229 representing the full 3d geometry of the channel we're able to pull out the full 21 00:01:31,229 --> 00:01:36,040 grain size distribution and actually see what's going on here 22 00:01:37,020 --> 00:01:37,880 we can then go 23 00:01:37,890 --> 00:01:42,150 through these different rivers to systems this is an animation of another 24 00:01:42,150 --> 00:01:46,380 3D model and we can move up and down stream throughout these channels and map 25 00:01:46,380 --> 00:01:49,979 just like we were in the field we can actually pull out things that are more 26 00:01:49,979 --> 00:01:53,130 difficult to map in the field this is actually one that we've had to survey 27 00:01:53,130 --> 00:01:57,090 from a trail it's not possible to get down into this Bank but we can go 28 00:01:57,090 --> 00:02:01,110 through and we can map lithology like we would in the field we can pull out full 29 00:02:01,110 --> 00:02:05,670 grain size distributions measuring every single grain that is in the channel we 30 00:02:05,670 --> 00:02:08,610 can pull out things that would be really difficult to measure in the field like 31 00:02:08,610 --> 00:02:12,330 the high watermarks that we're seeing in the channel breaks in that vegetation or 32 00:02:12,330 --> 00:02:15,299 that weathered bedrock we can also pull out where that current 33 00:02:15,300 --> 00:02:19,860 water is and then we can see how this changes at different points in the channe 34 00:02:22,420 --> 00:02:26,240 l so this is an example of the kind of data density that we can get out 35 00:02:26,240 --> 00:02:33,380 of this so this is a one kilometer reach in northeastern Taiwan and basically we 36 00:02:33,390 --> 00:02:38,670 have measurements of the bed slope across the entire channel the high-water 37 00:02:38,670 --> 00:02:43,590 marks the channel width at both high water and low flows. We can also measure 38 00:02:43,590 --> 00:02:48,030 the relative area of sediment both coarse sediment and fine sediment in 39 00:02:48,030 --> 00:02:53,639 Taiwan fine sediment is like less than a meter so different than other systems we 40 00:02:53,639 --> 00:02:57,420 can measure the distribution of boulders throughout the system and what we're 41 00:02:57,420 --> 00:03:02,310 actually starting to see here is that the spatial variability in these rivers 42 00:03:02,310 --> 00:03:06,959 is actually pretty much due to the rock strength so we have variations in the 43 00:03:06,959 --> 00:03:11,579 lithology of bedrock that is exposed at the surface and that is changing where 44 00:03:11,579 --> 00:03:16,290 you're getting wide rivers changing the channel morphology so here for example 45 00:03:16,290 --> 00:03:21,030 the black schist are relatively weakest rock is creating much wider river 46 00:03:21,030 --> 00:03:26,430 channels and that river channel it's actually widening when we are parallel 47 00:03:26,430 --> 00:03:30,389 to the foliation Ingle of the bedrock itself so you have structural 48 00:03:30,389 --> 00:03:35,129 variability that is really creating the spatial variability in the system and 49 00:03:35,129 --> 00:03:38,400 that's actually corresponding to where we're getting different sedimentation 50 00:03:38,400 --> 00:03:43,979 patterns so cool to see it change over space but we can also measure it over 51 00:03:43,979 --> 00:03:50,430 time so these are photorealistic surveys that have full 3d capability and so we 52 00:03:50,430 --> 00:03:55,169 can take a series of repeat surveys this is further downstream of the same river 53 00:03:55,169 --> 00:03:59,819 where we have for repeat surveys and we can actually compare them and see how 54 00:03:59,819 --> 00:04:04,319 these rivers change with different storm events so we have a really unique data 55 00:04:04,319 --> 00:04:10,229 set this is about a 200 meter reach in also in northeastern Taiwan where we 56 00:04:10,229 --> 00:04:14,250 have a range of grain sizes anywhere from kind of 12 meter boulders don't 57 00:04:14,250 --> 00:04:19,470 have kind of fine sand and these different surveys capture a range of 58 00:04:19,470 --> 00:04:24,930 different discharge events so between 2015 and 2017 we actually captured a 59 00:04:24,930 --> 00:04:28,990 Typhoon um between 2017 and 2018 we had a 60 00:04:28,990 --> 00:04:34,090 high-flow year not a typhoon but still high flow and then between 2018 and 2019 61 00:04:34,090 --> 00:04:39,310 there were relatively little storms so we can start to compare this kind of 62 00:04:39,310 --> 00:04:45,190 data so we can fully map across these reaches the different distributions of 63 00:04:45,190 --> 00:04:51,819 different grain sizes so this is between 2015 and 2017 that, that change we were 64 00:04:51,819 --> 00:04:56,620 capturing the typhoon event and we can actually measure the differences in the 65 00:04:56,620 --> 00:05:00,820 grain size distribution throughout so overall here we're not seeing a large 66 00:05:00,820 --> 00:05:07,060 change in the overall grain, grain size distribution but we are seeing a lot of 67 00:05:07,060 --> 00:05:11,770 change in 3D so we can actually measure aggradation and erosion at really high 68 00:05:11,770 --> 00:05:16,630 resolution and we're seeing meter up to a meter of aggradation and a meter of 69 00:05:16,630 --> 00:05:20,919 erosion throughout this system and it's non spatially uniform we can actually go 70 00:05:20,919 --> 00:05:23,949 through and measure every single boulder that was in this channel and see whether 71 00:05:23,949 --> 00:05:28,509 it moved or not so about half of the boulders in this system did move in this 72 00:05:28,509 --> 00:05:33,909 time period and while it is more likely for the finer grains the 1 meter 73 00:05:33,909 --> 00:05:38,320 boulders to move we are seeing up to an 8 meter boulder moving between these two 74 00:05:38,320 --> 00:05:46,599 on two surveys if we move in from the typhoon year to the stormy year we have 75 00:05:46,599 --> 00:05:50,469 less overall change we're seeing a net erosion throughout the system 76 00:05:50,469 --> 00:05:54,490 we're transporting fewer boulders but we're still transporting up to a 5 meter 77 00:05:54,490 --> 00:05:59,139 boulder in between that time but we're seeing a really big change in the grain 78 00:05:59,139 --> 00:06:03,099 size distribution that we're actually seeing in the channel so all the sudden 79 00:06:03,099 --> 00:06:08,020 we've lost a lot of our gravel and we're increasing all of our other size classes 80 00:06:08,020 --> 00:06:11,259 so we're actually making the channel more complex and we're reorganizing the 81 00:06:11,259 --> 00:06:14,919 channel and getting a lot more of a rougher surface that we're seeing there 82 00:06:14,919 --> 00:06:19,030 and this is kind of that gravel is eroding out in these patches all 83 00:06:19,030 --> 00:06:25,150 throughout the system between 2018 and 2019 where we have a relatively quiet 84 00:06:25,150 --> 00:06:29,949 year we don't see much change in the overall bed state but we still see more 85 00:06:29,949 --> 00:06:35,650 of that gravel going away and the other size class is increasing and so this 86 00:06:35,650 --> 00:06:40,920 kind of spatial pattern is showing that complex to move these kind of flows with 87 00:06:40,920 --> 00:06:45,000 this wide range of grain sizes but we can start to capture the full complexity 88 00:06:45,000 --> 00:06:50,540 both in the surface measurements and in the actual full grain size distributions 89 00:06:50,540 --> 00:06:54,960 so overall this has showed that we can use these kind of surveys to measure 90 00:06:54,960 --> 00:07:01,530 things over both space and over time and see how the different sediment how 91 00:07:01,530 --> 00:07:05,130 sediment is able to be transported differently wherever we are in the 92 00:07:05,130 --> 00:07:08,700 stream and we can use these repeat surveys to see how it'll change over 93 00:07:08,700 --> 00:07:11,840 time so thank you for having me here 94 00:07:12,480 --> 00:07:13,840 I’m done