Contribution of Excessive Supply of Solid Material to a Runoff

water

Case Report Contribution of Excessive Supply of Solid Material to a Runoff-Generated Debris Flow during Its Routing Along a Gully and Its Impact on the Downstream Village with Blockage Effects

Ming-liang Chen 1, Xing-nian Liu 1, Xie-kang Wang 1, Tao Zhao 2 and Jia-wen Zhou 1,*

1 State Key Laboratory of Hydraulics and Mountain River Engineering, Sichuan University, Chengdu 610065, China; [email protected] (M.C.); [email protected] (X.L.); [email protected] (X.W.) 2 College of Water Resource and Hydropower, Sichuan University, Chengdu 610065, China; [email protected] * Correspondence: [email protected]; Tel.: +86-28-8546-5055

Received: 3 January 2019; Accepted: 15 January 2019; Published: 18 January 2019

Abstract: On 8 August 2017, a runoff-generated debris flow occurred in the Puge County, Sichuan Province of southwestern China and caused huge property damage and casualties (25 people died and 5 people were injured). Emergency field investigations found that paddy fields, dry land, residential buildings and roads suffered different degrees of impact from the debris flow. This paper reveals the formation process of the debris flow by analyzing the characteristics of rainfall precipitation and sediment supply conditions in the study area and it approaches the practical application of hazard prevention and mitigation constructions. Doppler weather radar analysis indicates that a very high intensity rainfall occurred in the middle and upper zones of the basin, illustrating the importance of enhancing rainfall monitoring in high-altitude areas. The abundant supply of deposits in gully channels is among the significant causes of a transformation from mountain floods to large-scale debris flows. It was also found that the two culverts played an important role in the movement affecting the processes of debris flows which has substantially aggravated the destructive outcome. The excessive supply of solid material and local blockage with outburst along a gully must receive significant attention for the prediction of future debris flows, hazard prevention and mitigation measures.

Keywords: debris ﬂow; heavy rainfall; supply condition; scale ampliﬁcation; blockage with outburst

1. Introduction Debris flows have been posed a great threat to human beings, infrastructure (e.g., roads, tunnels, bridges) and residences in mountainous areas. In recent years, the number of debris flow events increased because of the combination of high intensity rainfalls and the availability of solid material due to the seismic load in Southwest China after the 2008 Wenchuan Earthquake [1–3]. Debris flow essentially is a solid-liquid mixture routing on very high slopes. More refined descriptions of debris flows are provided by Iverson [4,5]. A debris flow can be initiated from the shallow failure induced by high pore water pressure within shearing slip zones or from sediment entrainment into runoff by erosion. Runoff-generated debris flow is characterized by the rapid movement and the sufficient runoff to disperse solid grains throughout the whole depth of flow [6–8]. The movement process of runoff-generated debris flow is usually accompanied with the entrainment of a large quantity of debris material. Therefore, they occur in the mountainous areas of Southwest China when they are hit by heavy rainfall that can cause abundant runoff that is able to entrain large quantities of debris

Water 2019, 11, 169; doi:10.3390/w11010169 www.mdpi.com/journal/water Water 2019, 11, 169 2 of 17

Water 2019, 1 x FOR PEER REVIEW 2 of 17 material. Such a triggering mechanism, initially revealed by laboratory experiments on flumes [9] and slopesentrain [10 large–12], hasquantities also been of showndebris bymaterial. field investigations Such a triggering in all differentmechanism, environments initially revealed around by the worldlaboratory [13–20 experiments], including on China flumes [20 [9],21 and]. These slopes debris [10–12], flows, has also once been being shown triggered, by field can investigations dramatically growin all in different volume environments [22–29] due around to the entrainmentthe world [13– of20], debris including deposits China on [20,21]. the propagation These debris direction. flows, once being triggered, can dramatically grow in volume [22–29] due to the entrainment of debris The runout area and deposition depth depend on the debris flow volume [30–34]. Such events increased deposits on the propagation direction. The runout area and deposition depth depend on the debris over the past years due to the increase of (1) extreme rainfall [35], (2) the melting of glaciers [36] flow volume [30–34]. Such events increased over the past years due to the increase of (1) extreme induced by climate change and (3) in particular cases, great earthquakes, such as the Wenchuan rainfall [35], (2) the melting of glaciers [36] induced by climate change and (3) in particular cases, earthquakes [37–43]. great earthquakes, such as the Wenchuan earthquakes [37–43]. Around human settlements, prevention construction and measures for geological risks are Around human settlements, prevention construction and measures for geological risks are designed and established to ensure the safety of personnel and property and to reduce the impact designed and established to ensure the safety of personnel and property and to reduce the impact fromfrom geological geological hazards hazards toto thethe greatestgreatest extentextent possible.possible. Protection Protection and and stabilization stabilization structures, structures, sedimentsediment retention retention works works (e.g., (e.g., checkcheck dams,dams, sedimentsediment trapping dams) dams) and and siltation siltation and and discharge discharge controlcontrol works works (drainage (drainage channels, channels, grilledgrilled dams,dams, sedimentationsedimentation basins, basins, step-p step-poolool systems systems and and others) others) areare the the constructions constructions that that areare commonlycommonly usedused toto regulate the formation and and movement movement of of debris debris flowsflows [ 44[44].]. However,However, inin thethe mountainousmountainous areas of southwestern China, China, some some previous previous hazard hazard preventionprevention and and reduction reduction construction construction measuresmeasures diddid not effectively work work and and increased increased the the risk risk to to humanhuman beings beings [45 [45–47].–47]. One One of of the the mainmain reasonsreasons for thisthis is that that the the magnitude magnitude of of the the design design event event was was underestimatedunderestimated because because the the excess excess supply supply ofof solidsolid materialsmaterials was not considered. InIn this this work, work, a a debris debris flow flow that that occurred occurred in in the the Puge Puge County County will will be be used used as as a a case case study study (Figure (Figure1 ). Aimed1). Aimed at understanding at understanding the mechanism the mechanism of sediment of se supplydiment conditions supply conditions that influence that theinfluence evolution the of debrisevolution flows of and debris artificial flows and structures artificial that structures influence that movement influence movement process, the process, event the will event be analyzed will be inanalyzed detail in in the detail paper. in the Emergency paper. Emergency field investigations field investigations identified identified three three regions regions in the in studythe study area: thearea: formation the formation area, the area, transition the transition area and area the inundatedand the inundated or runout or area, runout which area, reflect, which respectively, reflect, therespectively, corresponding the different corresponding activity characteristicsdifferent activity of debris characteristics flows. After comprehensiveof debris flows. analyses After of thecomprehensive results of the fieldanalyses investigation of the results and three of essentialthe field factorsinvestigation (generation and ofthree surface essential runoff; factors supply condition(generation and of channel surface steepness), runoff; supply the contributionscondition and from channel the excessivesteepness), supply the contributions of solid materials from the and localexcessive blockage supply with of outburst solid materials along the and gully local channel blockage will with be outburst discussed along systematically the gully channel in this will paper. be discussed systematically in this paper.

FigureFigure 1. 1.An An overview overview of theof studythe study area byarea aerial by aerial photography photography on 8 August on 8 August 2017 with 2017 the with site location the site of thelocation study areaof the on study a map area of on China. a map of China.

Water 2019, 11, 169 3 of 17

Water 2019, 1 x FOR PEER REVIEW 3 of 17 2. Description of the Debris Flow Erosion Process Observed in the Tongzilin Gully 2. Description of the Debris Flow Erosion Process Observed in the Tongzilin Gully The study area belongs to the Hengduan Mountain range that runs east-west in the Yunnan-GuizhouThe study area Platea belongs and is locatedto the on Hengduan the east side Mountain of Luoji Mountain,range that approximately runs east-west 17 km in fromthe Yunnan-GuizhouPuge County (Figure Platea2). According and is located to the on data the of east the lastside three of Luoji years Mountain (2014–2017), approximately from meteorological 17 km fromstations Puge around County the study(Figure area, 2). the According annual average to the precipitation data of the islast 1716.9 three mm years with the(2014–2017) majority beingfrom meteorologicaldistributed over stations the period around from the May study to September. area, the Inannual August, average the mean precipitation monthly precipitationis 1716.9 mm of with the thestudy majority area is being 235 mm distributed and the maximumover the period daily from precipitation May to September. is 85.1 mm. In August, the mean monthly precipitation of the study area is 235 mm and the maximum daily precipitation is 85.1 mm.

Figure 2. GeologicGeologic and and geomorphic geomorphic map of the Zemu Zemuhehe fault zone around the study area.

Emergency field field investigations can help to understand the dynamic process and movement mechanism. Immediately Immediately af afterter the the debris debris flow, flow, an investigation team visited the site and collected geological information.information. As As shown shown in Table in Table1, the Tongzilin1, the Tongzilin Gully has aGully channel has length a channel of approximately length of 2 approximately5.5 km and a catchment5.5 km and areaa catchment of approximately area of approximately 3.7 km with 3.7 significant km2 with significant topographical topographical variation. variation.The study The area study elevation area overelevation 1650 over m is characterized1650 m is characterized by erosional by landformserosional landforms with a maximum with a maximumelevation ofelevation 3430 m, of an 3430 elevation m, an elevation difference difference of 1780 m,of 1780 a longitudinal m, a longitudinal gradient gradient of 476‰ of 476‰and a ◦ ◦ andslope a slope gradient gradient of 40 of–60 40°–60°.. The The section section below below (from (from the the elevation elevation of of 1650 1650 m m toto thethe gully bottom) is characterized characterized by by erosion erosion and and denudation denudation landforms landforms with with an an elevation elevation difference difference of of258 258 m, m,a ◦ ◦ longitudinala longitudinal gradient gradient of of 147‰ 147‰ andand a slope gradientgradient ofof 15 15°–20°.–20 . ToTo better better describe describe the the characteristics characteristics of ofthe the debris debris flow, flow, the the study study area area is divided is divided into into three three regions regions from from top totop bottom: to bottom: (1) the (1) formationthe formation area area(the elevation(the elevation over theover steep the cliff);steep (2)cliff); the transition(2) the transition area (from area the (from steep the cliff steep to the cliff township to the township road with roadarch culvert);with arch and culvert); (3) the and inundated (3) the orinundated runout area or runout (from thearea township (from the road township to the Zemuroad to River). the Zemu River). Table 1. Main morphometric parameters of the Tonzilin Gully.

Table 1. MainLocation morphometric parameters of the Tonzilin Gully. Tongzilin Gully Channel lengthLocation (km) Tongzilin 5.5 Gully 2 BasinChannel area (kmlength) (km) 3.7 5.5 Average longitudinal channel gradient (‰) 365 2 Average longitudinal channelBasin gradient area (Elevation (km ) 1650–3430 m, ‰) 476 3.7 Average longitudinalAverage channel longitudinal gradient channel (Below elevation gradient 1650 (‰) m, ‰) 147 365 AverageAverage longitudinal slope gradient channel (Elevation gradient 1650–3430 (Elevation m, ‰)1650–3430 m, ‰) 40◦–60◦ 476 ◦ ◦ AverageAverage longitudinal slope gradient channel (Below gradient elevation (Below 1650 elevation m, ‰) 1650 m, ‰) 15 –20 147 Average slope gradient (Elevation 1650–3430 m, ‰) 40°–60° Two branchAverage catchment slope gradient channels (Below occur elevation in the formation 1650 m, area‰) before convergence, 15°–20° of which the flow directions of the catchment channels are N21◦E and N73◦E, respectively. After the convergence, the flowTwo directionbranch catchment of the catchment channels channel occur in turns the formation out to be N61area◦ E.before Little convergence, attention was of paid which to the flowcatchment directions channel of the with catchment a flow direction channels of are N21 N21°◦EE in and the formationN73°E, respectively. area and the After part the with convergence, a distance the flow direction of the catchment channel turns out to be N61°E. Little attention was paid to the catchment channel with a flow direction of N21°E in the formation area and the part with a distance over 400 m from the convergence in the catchment channel with a flow direction of N73°E due to the

Water 2019, 11, 169 4 of 17

WaterWater 2019 2019, 1 ,x 1 FOR x FOR PEER PEER REVIEW REVIEW 4 of4 of17 17 over 400 m from the convergence in the catchment channel with a flow direction of N73◦E due to difficultiesthedifficulties difficulties raised raised raised by by the bythe steepness. the steepness. steepness. The The right Theright rightcha channelnnel channel in in the the information formation the formation area area is area ischaracterized characterized is characterized by by an an irregularbyirregular an irregular topography topography topography transforming transforming transforming from from a V-shapea from V-shape a V-shapeto to a U-shapea U-shape to a with U-shape with a beda bed with width width a bedranging ranging width between between ranging 5 5 andbetweenand 12 12 m. 5m. andAs As 12shown shown m. As in shownin Figure Figure in Figure3, 3,the the 3slope, theslope slopegradie gradie gradientntnt of of channel ofchannel channel banks banks banks of of ofthe the Tongzilin TongzilinTongzilin gully gully gully isis is approximatelyapproximately 2525°–35°. ◦25°–35°.–35◦. TheThe The slope slope slope is mainlyis ismainly mainly covered covered covered by Quaternary by by Quaternary Quaternary alluvium, alluvium, deluviumalluvium, deluvium anddeluvium colluvium, and and colluvium,withcolluvium, several with bedrockswith several several exposed bedrocks bedrocks locally exposed inexposed steep slopeslocally locally and in in cliffs.steep steep The slopes slopes bedrock and and is characterizedcliffs. cliffs. The The bedrock bybedrock purple is is characterizedsiltstonecharacterized and mudstoneby by purple purple siltstone with siltstone a dip and and angle mudstone mudstone of approximately with with a adip dip angle 40angle◦. of Because of approximately approximately of the influence 40°. 40°. Because Because of fault of of theactivities,the influence influence the rockof of fault massfault activities, isactivities, broken the with the rock abundantrock mass mass is joints isbroken broken and fractures.with with abundant abundant As shown joints joints in and Figure and fractures. 4fractures., the slope As As shownsurfacesshown in ofinFigure Figure both 4, the 4,the channelthe slope slope surfaces banks surfaces have of of both aboth sharply th the channele channel scoured banks banks topography, have have a sharplya sharply with thescoured scoured loss of topography, vegetationtopography, withcover.with the Thethe loss loss toes of of ofvegetation vegetation the banks cover. havecover. beenThe The toes incised toes of of the and the banks erodedbanks have have with been anbeen undercuttingincised incised and and eroded depth eroded ofwith with 1–6 an m. an undercuttingAundercutting number of outcropsdepth depth of of 1–6 are 1–6 m. characterized m. A Anumber number of by of outcrops rocks outcrops wrapped are are characterized characterized in sediment by by and rocks rocks a fewwrapped wrapped plant rootsin in sediment sediment on the andundercuttingand a fewa few plant plant surface. roots roots on A on significantthe the undercutting undercutting amount surface. ofsurface. residual A Asignificant materialssignificant amount were amount retained of of residual residual in the materials gullymaterials channel, were were retainedmostretained of whichin in the the are gully gully washed channel, channel, rocks most characterizedmost of of which which byare are siltstone washed washed (purple, rocks rocks characterized grayishcharacterized green by andby siltstone siltstone grayish (purple, yellow). (purple, grayishThesegrayish rocks green green were and and transportedgrayish grayish yellow). yellow). by the These debrisThese rocks rocks flow were fromwere transported placestransported with by higher by the the debris elevations.debris flow flow from from places places with with higherhigherThe elevations. elevations. local side edges are covered by several rocks in the gully bed that are wrapped in sediments and originatedTheThe local local side fromside edges edges the collapseare are covered covered of the by by bankseveral several slope. rocks rocks The in in the residual the gully gully crushedbed bed that that are rocks are wrapped wrapped in the gully in in sediments sediments channel andvaryand originated in originated size, mainly from from fromthe the collapse 0.05 collapse m toof 0.15of the the m,bank bank and slop aslop fewe. e.The largeThe residual residual stones crushed vary crushed from rocks rocks 1.0–2.5 in in the m. the gully gully channel channel varyvary in in size, size, mainly mainly from from 0.05 0.05 m m to to 0.15 0.15 m, m, and and a fewa few large large stones stones vary vary from from 1.0–2.5 1.0–2.5 m. m.

(a()a ) (b()b )

FigureFigure 3.3. (3.a(a) )( An aAn) overviewAn overview overview of theof of formationthe the formation formation area inarea thearea studyin inthe the area study study by aerialarea area photography.by by aerial aerial photography. photography. (b) Lithological (b ()b ) LithologicalcompositionLithological composition exposed composition behind expo expo thesedsed bank behind behind slope. the the bank bank slope. slope.

(a()a ) (b()b )

Figure 4. (a) An overview of the gully channel in the formation area. (b) Morphological observation of FiguretheFigure residual 4. 4.(a )( a materialAn) An overview overview in the of side ofthe the edge gully gully of ch the channelannel gully in in channelthe the formation formation of the formationarea. area. (b ()b Morphological) area.Morphological observation observation of ofthe the residual residual material material in inthe the side side edge edge of ofthe the gully gully channel channel of ofthe the formation formation area. area.

TheThe transition transition area area flows flows on on a steepa steep cliff cliff and and a gentlera gentler slope. slope. As As shown shown in in Figure Figure 5a, 5a, a largea large bulk bulk ofof bedrock bedrock was was exposed exposed at at the the top top region region of of the the steep steep cliff. cliff. The The altitude altitude difference difference of of the the steep steep cliff cliff is is

Water 2019, 11, 169 5 of 17

The transition area flows on a steep cliff and a gentler slope. As shown in Figure5a, a large bulk of bedrock was exposed at the top region of the steep cliff. The altitude difference of the steep cliff is approximately 200 m. In the steep cliff, the scale amplification of the debris flow occurred due to the erosion of the accumulated deposits in the gully channel and its movement also accelerated substantially because of the large altitude difference which led to a change of morphological characteristics similar to that in the formation area. During the flows, the debris interrupted the Water 2019, 1 x FOR PEER REVIEW 5 of 17 mountain road below the exposed bedrock and destructed the roadbed at the junction of the steep cliff andapproximately gentle slope. 200 Because m. In the of steep the cl hydrauliciff, the scale erosion amplification on the of th sedimente debris flow bed, occurred the surface due to runoffthe took away looseerosion solid of materialthe accumulated and the deposits shallow in surfacethe gully ofchannel accumulated and its movement deposits also and accelerated gradually turned substantially because of the large altitude difference which led to a change of morphological into a debris flow [9,48,49]. Field surveys on the surface geomorphology after the debris flow event characteristics similar to that in the formation area. During the flows, the debris interrupted the show thatmountain a layer road of mud below sediment the exposed was bedrock left andand destru accumulatedcted the roadbed on the at scoured the junction surface of the onsteep both sides. Three cross-sectionscliff and gentle are slope. selected Because from of the downstream hydraulic erosion to upstream on the sediment for the bed, measurement the surface runoff of morphological took characteristicsaway loose (Table solid2). material Because and of the the shallow great surface erosive of energy accumulated caused deposits by a and large gradually slope gradient,turned both channel banksinto a debris were flow eroded [9,48,49]. and Field incised surveys with on th ane surface undercutting geomorphology depth after of approximately the debris flow event 1 m. This is show that a layer of mud sediment was left and accumulated on the scoured surface on both sides. particularly evident at the location near the corner where the sediment deposits have a larger covering Three cross-sections are selected from downstream to upstream for the measurement of width. Inmorphological the channel characteristics and its neighboring (Table 2). area,Because broken of the great stones erosive of different energy caused sizes by are a leftlarge on slope the earthen surface bygradient, transportation both channel from banks higher were eroded places. and incised with an undercutting depth of approximately 1 m. This is particularly evident at the location near the corner where the sediment deposits have a Tablelarger 2. Morphological covering width. characteristics In the channelof and three its neig chosenhboring sections area, afterbroken suffering stones of the different influence sizes of are debris flowleft in theon the transition earthen area.surface by transportation from higher places.

Table 2. Morphological characteristics of three chosen sections after suffering the influence of debris Left Bank Right Bank Riverbed Water flow in the transition area. No. Longitude Latitude Covering Covering Width Passing Object Object Left Bank Width (m) Right Bank Width (m) (m) Height (m) Riverbed Water Passing No. ◦Longitude0 00 Latitude◦ 0 00 Covering Covering 1 E 102 28 37 N 27 28 09 PaddyObject ﬁeld 34.8Object Dry land 8.0Width (m) 7.3Height (m) 0.9 Width (m) Width (m) 2 E 102◦2803500 N 27◦2800700 Woods 12.0 Dry land 15.7 9.0 1.0 1 E 102°28′37′′ N 27°28′09′′ Paddy field 34.8 Dry land 8.0 7.3 0.9 3 E 102◦2802900 N 27◦2705800 Hennery 14.5 Barren slope 5.2 10.5 0.8 2 E 102°28′35′′ N 27°28′07′′ Woods 12.0 Dry land 15.7 9.0 1.0 3 E 102°28′29′′ N 27°27′58′′ Hennery 14.5 Barren slope 5.2 10.5 0.8

(a)

(b)

Figure 5. Cont.

Water 2019, 11, 169 6 of 17 Water 2019, 1 x FOR PEER REVIEW 6 of 17

(c)

(d)

Figure 5. (a) A photo of the gully channel between the formation and transition areas. (b) A photo of Figurethe 5. gully(a) A channel photo inof the the transition gully channel area. (c )between Scour of thethe right formation bank of and the gully transition channel areas. in the ( transitionb) A photo of the gullyarea. channel (d) Scoring in ofth thee transition left bank of area. the gully (c) channelScour of in the the transitionright bank area of (all the photos gully were channel shot on in the transition9 August area. 2017). (d) Scoring of the left bank of the gully channel in the transition area (all photos were shot on 9 August 2017). Two roads cross the channel in the inundated or runout area at elevations of 1505 m and 1420 m, Tworespectively. roads cross The township the channel road in was the constructed inundated in 2015or runout at an elevation area at elevations of 1505 m. In of addition, 1505 m aand ﬂat 1420 section of channel with a width of 8–10 m was formed because of excavation and an arch culvert was m, respectively. The township road was constructed in 2015 at an elevation of 1505 m. In addition, a built up to discharge the surface runoff around the subgrade. The arch culvert has a width of 3.0 m, flat sectiona sidewall of channel height of with 1.5 m a andwidth an archof 8–10 height m ofwas 0.8 formed m (Figure because6). As shown of excavation in Figure 6and, a large an arch stone, culvert was builtwith up an approximateto discharge length the surface of 4.7 m, runoff a width around of 2.0 m the and subgrade. a height of The 1.7m, arch was culvert found whenhas a cleaning width of 3.0 m, a sidewallthe culvert height after the of events.1.5 m and The giantan arch stone height that was of 0.8 transported m (Figure by debris6). As ﬂowshown blocked in Figure the culvert 6, a large stone,with with the an expansion approximate of the length inundated of 4.7 or runoutm, a width area by of accumulating 2.0 m and a solidheight material of 1.7 atm, the was entrance found of when cleaningthe archthe culvert culvert, after further the causing events. discharge The giant diversion stone that because was of transported the burst after by blockage. debris flow blocked the culvert with the expansion of the inundated or runout area by accumulating solid material at the entrance of the arch culvert, further causing discharge diversion because of the burst after blockage.

Water 2019, 1 x FOR PEER REVIEW 7 of 17 Water 2019, 11, 169 7 of 17 Water 2019, 1 x FOR PEER REVIEW 7 of 17

Figure 6. Route diversion of the debris flow after the arc culvert of the township. road in the Figureinundated 6. Route or runout diversion area ofwith the morphological debris flow after observ the arcation culvert of the of cleared the township arc culvert road inof thethe inundatedtownship Figure 6. Route diversion of the debris flow after the arc culvert of the township road in the orroad runout and areathe great with morphologicalstone block in observationthe arc culvert of the after cleared cleaning arcculvert (all photos of the were township shot on road 9 andAugust the inundated or runout area with morphological observation of the cleared arc culvert of the township great2017). stone block in the arc culvert after cleaning (all photos were shot on 9 August 2017). road and the great stone block in the arc culvert after cleaning (all photos were shot on 9 August After2017). rapidly changing the flowflow direction, the debrisdebris flowflow spread and inundated the #2 and #3 residential quarters,quarters, causingcausing 18 18 deaths deaths and and the the destruction destruction of of houses. houses. The The sediment sediment deposits deposits silted silted up After rapidly changing the flow direction, the debris flow spread and inundated the #2 and #3 inup front in front of the of buildingsthe buildings had ahad covering a covering height height of approximately of approximately 3–4 m. 3–4 A smallm. A portionsmall portion of the debrisof the residential quarters, causing 18 deaths and the destruction of houses. The sediment deposits silted flowdebris ran flow through ran through the culvert the culvert or over or the over road the surface road andsurface continued and continued to run along to run the along original the channel.original up in front of the buildings had a covering height of approximately 3–4 m. A small portion of the Thechannel. village The road village was road constructed was constructed in 2014 atin an2014 elevation at an elevation of 1420 of m. 1420 A rectangular m. A rectangular culvert culvert had a debris flow ran through the culvert or over the road surface and continued to run along the original heighthad a height of 1.5 mof and1.5 m a widthand a ofwidth 2.6 m of where 2.6 m the where original the original channel channel has a bend haswith a bend a turning with a angleturning of channel. The village road was constructed in 2014 at an elevation of 1420 m. A rectangular culvert approximatelyangle of approximately 60◦. As shown 60°. As in Figureshown7 in, when Figure running 7, when through running this through area, the this debris area, flow the carried debris large flow had a height of 1.5 m and a width of 2.6 m where the original channel has a bend with a turning blockcarried stones, large occludingblock stones, the rectangular occluding culvertthe rectan andgular further culvert causing and the further second causing discharge the diversion. second angle of approximately 60°. As shown in Figure 7, when running through this area, the debris flow Adischarge flow ran diversion along the. original A flow channelran along into the the original Zemu Riverchannel and into the the other Zemu flow River rushed and straight the other down flow to carried large block stones, occluding the rectangular culvert and further causing the second therushed residential straight quarters down to along the residential the village roadquarters killing along seven the people village and road covering killing theseven buildings people withand discharge diversion. A flow ran along the original channel into the Zemu River and the other flow sediments.covering the Eventually, buildings the with two se flowsdiments. entered Eventually, the Zemu the River two and flows turned entered into twothe pilesZemu of River sediments and rushed straight down to the residential quarters along the village road killing seven people and andturned blocked into two the piles river of channel sediments with and a shape blocked of a the circular river sectorchannel in with a view a shape of a vertical of a circular angle. sector in a viewcovering of a verticalthe buildings angle. with sediments. Eventually, the two flows entered the Zemu River and turned into two piles of sediments and blocked the river channel with a shape of a circular sector in a view of a vertical angle.

(a) (b)

Figure 7. (a) FlowFlow directiondirection(a) diversiondiversion of of the the debris debris ﬂow flow after after the the rectangle rectangle culvert culvert(b) of of the the village village road road in thein the inundated inundated or runoutor runout area. area. (b) ( Hazardb) Hazard trace trace situation situation after after the the diversion diversion of the of the debris debris ﬂow flow in the in inundatedtheFigure inundated 7. (a or) Flow runout or runout direction area area (all diversion photos(all photos were of werethe shot debris shot on 9 onflow August 9 Augustafter 2017). the 2017). rectangle culvert of the village road in the inundated or runout area. (b) Hazard trace situation after the diversion of the debris flow in 3. Formation the inundated and Evolutionor runout area Mechanisms (all photos were shot on 9 August 2017). 3.1. Meteorological Analyses 3.1.3. Formation Meteorological and AnalysesEvolution Mechanisms The closest meteorological station is located at the northwest of the study area and the distance is The closest meteorological station is located at the northwest of the study area and the distance about3.1. Meteorological 1.2 km from Analyses the study area. Rainfall data are shown at hourly intervals in Figure8a, together is about 1.2 km from the study area. Rainfall data are shown at hourly intervals in Figure 8a, The closest meteorological station is located at the northwest of the study area and the distance is about 1.2 km from the study area. Rainfall data are shown at hourly intervals in Figure 8a,

Water 2019, 11, 169 8 of 17

Water 2019, 1 x FOR PEER REVIEW 8 of 17 with the cumulative rainfall. Rainfall intensity progressively increased and reached a maximum togethervalue of with 13.5 mm/hthe cumulative between 3:00 rainfall. and 4:00 Rainfall a.m. andintensity decreased progressively thereafter. increased and reached a maximum value of 13.5 mm/h between 3:00 and 4:00 a.m. and decreased thereafter.

(a) (b)

FigureFigure 8. 8. RainfallRainfall characteristi characteristicscs on on 8 8August August 2017. 2017. (a ()a Hourly) Hourly and and cumulative cumulative rainfall rainfall measured measured by by thethe Qiaowo Qiaowo meteorological meteorological station.station. (b(b)) Evolution Evolution of of the the reﬂectivity reflectivity factor factor at an at elevation an elevation of 2.4 degreesof 2.4 degreeson the Doppleron the Doppler radar of radar Xichang of Xichang city. city.

TheThe study study area area was was completely completely covered covered by by a aDoppler Doppler weather weather radar radar which which is is located located at at Xichang Xichang citycity and and can can cover cover an anarea area with with a radius a radius of 150 of km. 150 km.The radar The radarlocation location is approximately is approximately 47 km away 47 km ◦ fromaway the from study the area study and areathe scan and elevation the scan of elevation the reflectivity of the reflectivityfactor is 2.4°. factor As shown is 2.4 in. Figure As shown 8b, at in approximatelyFigure8b, at approximately 2:02 a.m., a small 2:02 part a.m., of alocal small convection part of local with convection a peak echo with intensity a peak of echo 30 dBz intensity (unit of of radar30 dBz reflection (unit of factor) radar was reflection observed factor) in the was northw observedest of the in thestudy northwest area. Then, of at the approximately study area. Then,2:34 a.m.,at approximately the peak of 2:34 the a.m., echo the intensity peak of the was echo obse intensityrved wasin correspondence observed in correspondence with regions with located regions southwardslocated southwards and the peak and theecho peak intensity echo intensity was increased. was increased. At approximately At approximately 3:31 a.m., 3:31 the a.m., main the echo main wasecho observed was observed in the immediate in the immediate vicinity of vicinity the stud ofy the area; study combined area; combinedwith local withconvection, local convection, it caused theit causedinitiation the of initiation sustained of precipitation. sustained precipitation. In this moment, In this the moment, peak echo the peak intensity echo (reaching intensity (reaching40 dBz) was40 dBz)located was in located the study in the area. study After area. 4:03 After a.m., 4:03 a.m.,the echo the echointensity intensity in the in thestudy study area area decreased, decreased, howeverhowever the the peak peak intensity intensity of of the main echoecho waswas stillstill 3030 dBz. dBz. After After that, that, the the peak peak echo echo intensity intensity in in the thestudy study area area weakened. weakened. AccordingAccording to to the the dynamic analysisanalysis of of the the Doppler Doppler weather weather radar, radar, the the occurrence occurrence time time of the of debris the debrisflow wasflow estimated was estimated to fall to between fall between 3:30 a.m. 3:30 and a.m. 4:00 and a.m. 4:00 In a.m. the previousIn the previous hazards, hazards, such as such the Zhouqu, as the Zhouqu,Wenjiagou Wenjiagou and Yingxiu and Yingxiu events, theeven cumulativets, the cumulative precipitation precipitation before thebefore hazard the hazard occurrences occurrences was not wassmaller not smaller than 100 than mm 100 [37 mm–39 ].[37–39]. As shown As shown in Figure in Figure8, the recorded8, the recorded cumulative cumulative precipitation precipitation in this inevent this event was only was 49.2 only mm, 49.2 which mm, which is a relatively is a relatively smallvalue. small Thevalue. elevation The elevation difference difference of the basinof the in basinthe Tongzinlin in the Tongzinlin gully is approximatelygully is approximately 2038 m and 2038 the m basin and areasthe basin above areas 2000 above m are the2000 regional m are highthe regionalintensity high rainfall intensity zone withoutrainfall azone meteorological without a me station.teorological The current station. meteorological The current stationmeteorological is located stationat an elevationis located of approximatelyat an elevation 1400 of m.approximat Therefore,ely the 1400 measured m. Therefore, value of the precipitation measured is value likely lessof precipitationthan the real is value likely in less the than middle the andreal uppervalue in reaches the middle of the and basin upper area. reaches of the basin area.

3.2.3.2. Formation Formation Process Process and and Excessive Excessive Supply Supply of of Solid Solid Materials Materials TheThe supply supply of of solid solid material material is is a avery very important important factor factor for for the the dynamic dynamic evolution evolution of of mountain mountain hazardshazards [20,25,27,28,41,48]. [20,25,27,28,41,48]. According toto debris debris flow flow trace trace characteristics, characteristics, the excessivethe excessive supply, supply, coming comingfrom accumulated from accumulated deposits, deposits, granular granular solid materials solid materials and falling and falling rocks in rocks the Tongzilinin the Tongzilin gully, is gully, found isto found be a to significant be a significant factor forfactor the for continuous the continuo scaleus scale amplification amplification of a runoff-generatedof a runoff-generated debris debris flow flowduring during the formationthe formation and and routing routing processes. processes. Scouring, Scouring, undercutting undercutting and causingand causing collapse collapse are major are majormethods methods of obtaining of obtaining supply supply from solid from materials solid ma forterials runoff-generated for runoff-generated debris flows. debris The flows. formation The formationarea is characterized area is characterized by the narrow by the breach narrow width breach of the width channel, of the the channel, high steepness the high of steepness two side of channel two sidebanks channel and thebanks large and gradients the large of gradients the slopes. of Thesethe slopes. characteristics These characteristics are supposed are to supposed be favorable to be to favorable to the generation and development of debris flow [15]. The sudden occurrence of strong surface runoff caused by high intensity rainfall have a great erosive power of runoff. Accumulated

Water 2019, 11, 169 9 of 17 the generation and development of debris flow [15]. The sudden occurrence of strong surface runoff caused by high intensity rainfall have a great erosive power of runoff. Accumulated deposits in the toe region of the slope bank were distributed along the flow direction in the gully channel. Because of the great erosive energy, the toe region was scraped and eroded for the supplement to debris flow. Meanwhile, it has caused local slope failures and collapses in two sides of channel bank. These collapses produced many loose deposits, most of which were carried off by flow action; the rest were retained in the side edge of the gully bed. As shown in Figures4 and5, the cascading scale amplification, caused by the excessive supply of solid materials in the formation and transition areas, increases the influencing range of the debris flow and eventually generates serious destruction in the inundated or runout area [31,32,34]. On the bed of the gully, some debris deposited during the debris flow routing is present. In conclusion, the excessive supply of solid material for debris flow played an important role in changing the morphological appearance, lithologic proportion of accumulated formation and discharge capacity of the Tongzilin gully.

3.3. Affecting Process and Local Blockage with Outburst In the content of geomorphology, the outburst refers to the sudden release of surface runoff or debris flow when the barrier collapses or is eroded. The local blockage and outburst along the gully channel caused by culverts of the road project is another reason for the painful outcome. That is because the blockage and outburst changes the route of the debris flow which may aggravate the property damage and casualties. However, the original plan of facilitating the siltation and discharge control eventually enhanced the geologic hazard due to the fact that the culvert designs underestimated the discharge of the debris flow. The blockage of the culvert is a significant phenomenon in debris flow [27]. When the debris flow moved to the arch culvert (as shown in Figure6), the flow movement was obstructed because of improper design. At that moment, the arch culvert and the solid materials inside played the role of a temporary dam and the part of surface runoff flow along the original channel by overflow. With the continuous supplement from the debris flow, the solid materials were collected behind the temporary dam as well as on the road surface nearby. During the accumulation process of solid materials, the accumulation of solid materials above the road surface had a shocking impact until they were destroyed. When the part of the blockage above the road surface was destroyed, the accumulated solid materials generated an outburst. After the destruction, the debris flow changed its direction with a tendency of running to the lower elevation since the road surface is characterized by a dip angle from the horizontal level. Eventually, the debris flow changed flow direction and inundated at the #2 and #3 residential quarters with tremendous velocity. If no blockage effect occurred within the arch culvert, only the #1 residential quarter may have been inundated and the damage may have been substantially reduced. However, in this event, 18 deaths were caused in the #2 and #3 residential quarters. The blockage secondly occurred in the rectangle culvert, however with several differences from the first one because of topographic differences: (i) the original channel has a turn of 60◦ in the rectangle culvert; (ii) the road surface is parallel to the horizontal level. Known from the post-disaster situation, the debris flow was divided into two parts when passing through the second culvert. The original channel has a change of direction after the culvert. One part ran along the original channel by overflow and the other part continued the preceding flow direction by outburst.

3.4. Motion Characteristics According to the specification of geological investigation of debris flow stabilization (DZ/T0220-2006), published by the China Ministry of Lands and Resources, velocity and discharge of debris flow can be calculated by the following equations:

2 1 3 Vc = KcHc I 5 (1) Water 2019, 11, 169 10 of 17

Q = WcVc (2) Water 2019, 1 x FOR PEER REVIEW c 10 of 17 where: VKc is thea factor peak that velocity is related of the to debris debris flow flow (m/s); depth of which the value can be obtained by Figure 9; Kc is a factor that is related to debris flow depth of which the value can be obtained by Figure9; c Hc is the debris-flowdebris-flow depth depth determined determined by by the the measurement measurement of mudof mud trace trace at the at crossthe cross sections sections (m); I(m); is the longitudinal slope gradient of the channel (‰); 3 QI isc isthe the longitudinal peak discharge slope of gradient the debris of flowthe channel (m /s); (‰); c 2 3 WQ c isis the the peak cross-sectional discharge of area the (m debris). flow (m /s); Wc is the cross-sectional area (m2).

Figure 9. Recommended values of the velocity coefﬁcientcoefficient (K(Kcc) by debris ﬂowflow depth (Hcc).

The validity of above equations has been verifi veriﬁeded [38,50,51] [38,50,51] According to to geological geological survey, survey, dimensions ofof three three chosen chosen sections sections can becan obtained be obta byined Table by2 ,Table where 2, the where longitudinal the longitudinal slope gradient slope of thegradient channel of the in thesechannel places in these is 229‰. places The is estimated229‰. The results estimated are shown results in are Table shown3. in Table 3.

Table 3. The calculation results of motion characteristics’ parameters in three chosen sections. Table 3. The calculation results of motion characteristics’ parameters in three chosen sections.

2 Chinese Speciﬁcation No. KC I (‰) WC (m ) Chinese Specification No. KC I (‰) WC (m2) 3 3 V (m/s) QCC (m //s)s) VC C (m/s) 1 110 10229 229 25.83 25.83 179.3 6.9 6.9 2 210 10229 229 22.85 22.85 170.2 7.4 7.4 3 10 229 16.28 104.5 6.4 3 10 229 16.28 104.5 6.4

As shown in Table4 4,, althoughalthough thethe TongzilinTongzilin Gully Gully in in the the PugePuge eventevent hashas aa smallersmaller basinbasin withwith aa 2 drainage area of 3.7 kmkm2, the estimated values of peak velocity and peakpeak dischargedischarge are consideredconsidered as having a high levellevel whenwhen thethe accumulatedaccumulated precipitationprecipitation is 49.249.2 mm.mm. Although the dimension of the basin area limited the scale of debris flow, flow, thethe flowflow inin thethe PugePuge eventevent isis consideredconsidered as havinghaving a surprisingly large velocity when accessing residentialresidential quarters. For the prevention and mitigation of the geologic hazards, the most importantimportant thing isis toto accuratelyaccurately estimateestimate thethe motionmotion characteristiccharacteristic parameters of the potentialpotential debrisdebris flow.flow. The The Pu Pugege event has provided a lesson that erroneous estimations cannot reduce, however th theyey can instead increase the catastrophic impact of the debris flow.flow. InIn this this situation, situation, the the calculation calculation results results are shownare shown in Table in 4Table, which 4, provides which provides a design referencea design forreference the reconstruction for the reconstruction of flood discharge of flood discharge structures. structures.

Table 4. The comparison of motion characteristics of a section in the transition area between the Puge Table 4. The comparison of motion characteristics of a section in the transition area between the event and the Zhouqu event. Puge event and the Zhouqu event. Channel Basin Area LocationLocation Torrent Torrent Channel Length (km) Basin Area (kmV 2(m/s)) VC (m/s)Q (mQ3/s)C (m3/s) Length (km) (km2) C C Sanyanyu Gully 10.4 25.75 9.7 1358 Zhouqu [22] Sanyanyu Gully 10.4 25.75 9.7 1358 Zhouqu [22Luojiayu] Gully 9.5 16.14 11.0 572 Luojiayu Gully 9.5 16.14 11.0 572 Puge PugeTongzilin Tongzilin Gully Gully 5.5 5.5 3.7 3.7 6.9 6.9 105 105

4. Discussion The formation of a debris flow requires a large volume of surface runoff for the triggering of such a large-scale event [5,11,27–29]. In this case, the recorded cumulative rainfall did not provide a

Water 2019, 11, 169 11 of 17

4. Discussion The formation of a debris flow requires a large volume of surface runoff for the triggering of Watersuch 2019 a large-scale, 1 x FOR PEER event REVIEW [5,11 ,27–29]. In this case, the recorded cumulative rainfall did not provide11 of 17 a runoff volume that was large enough for triggering such a large-scale event. It is well-recognized runoffin meteorology volume that that was when large the enough air current for triggering suffers obstruction such a large-scale by high event. mountains, It is well-recognized large amounts in of meteorologylocal rainfall that are very when likely the air to occurcurrent in suffers places obstru of highction elevation. by high Therefore, mountains, it can large be amounts inferred thatof local the rainfalllocal heavy are very rainfall likely was to generated occur in places in the upstreamof high elevation. region of Therefore, the Tongzilin it can gully. be inferred The event that provides the local a heavywarning rainfall of the was need generated for a complete in the andupstream responsive region rainfall of the monitoringTongzilin gully. system The in theevent mountainous provides a warningareas of Southwestof the need China. for a complete and responsive rainfall monitoring system in the mountainous areasAccording of Southwest to local China. residents, a large-scale mountain flood has not occurred in the past 100 years. WhenAccording a rainfall to with local intermediate residents, intensitya large-scale occurs, mountain a small-scale flood surfacehas not runoffoccurred is formed in the withpast low100 years.scraped When depth. a rainfall In terms with of intermediate the impact ofintensity the small-scale occurs, a surfacesmall-scale runoff surface on the runoff supply is formed conditions, with lowabundant scraped accumulated depth. In terms deposits of theexist impact in of the the gully small-scale channel surface that originated runoff on fromthe supply mass conditions, movement abundantprocesses (shallowaccumulated landslides, deposits collapses, exist in rockfallthe gull andy channel others) that and originated tectonic activities. from mass Due movement to the rare processesoccurrence (shallow of severe landslid mountaines, collapses, floods in rockfall the past, and plenty others) of solid and materialtectonic hasactivities. remained Due into the the gully rare occurrencechannel in theof severe past 100 mountain years. Different floods in discharges the past, pl generateenty of differentsolid material scale ofhas debris remained flow in as the matter gully of channelthe availability in the past of loose 100 years. sediment Different at the channeldischarges bed ge [27nerate–29]. different In this event, scale the of debris abundant flow surface as matter runoff of thecarried availability out more of suppliesloose sediment of solid at materials the channel from be thed [27–29]. actions ofIn scour,this event, undercutting the abundant and collapse.surface runoffThe abundant carried surfaceout more runoff supplies is able of to solid mobilize material larges quantitiesfrom the actions of sediments of scour, with undercutting the flow velocity and collapse.depending The on abundant runoff discharge, surface runoff while is flow able depth to mo andbilize sediment large quantities concentration of sediments is dependent with the on flow the velocitygrain size depending [52,53]. The on constantrunoff discharge, supplies while persistently flow depth improve and thesediment motion concentration features (velocity, is dependent density onand the depth) grain ofsize flow [52,53]. and furtherThe constant persistently supplies enhance persistently the erosion improve action. the moti Ason shown features in Figure(velocity, 10, densitythis interaction and depth) generates of flow the and scale further amplification persistently and energyenhance escalation the erosion to develop action. As flows shown in a cascadingin Figure 10,way. this Eventually, interaction the generates excessive the supply scale of amplification solid material and allows energy the developmentescalation to develop of a large flows scale in and a cascadingcatastrophic way. debris Eventually, flow. the excessive supply of solid material allows the development of a large scale and catastrophic debris flow.

Figure 10. Evolution of mountain water hazards and th thee cascading process process of the supply for hazard formationformation and and development. development.

The dynamics of of the propagating surge is a fu functionnction of different parameters, including the sedimentsediment content, content, channel channel slope slope angle angle and and run runoff [54]. off [As54]. shown As shown in Figure in Figure11, the common11, the common types of naturaltypes of topography natural topography that are favorable that are for favorable the blocka forge the of blockage debris flows of debris included ﬂows (i) included the constricted (i) the sectionconstricted of gully section channel, of gully (ii) arc channel, gully channels (ii) arc gully and channels(iii) the intersections and (iii) the between intersections the gully between channels the of different longitudinal gradients. The causes of the blockage were the reduction of flow area, the change of flow type and the decrease of flow velocity. The occurrence of the debris outburst, however, requires some necessary conditions, i.e., certain slope gradients on both sides and sufficient supply of solid materials. The slope in this area is sufficiently steep for the generation of large scale surface runoff, implying a significant amount of gravitational energy. As such, the gravitational energy of abundant solid materials in the debris flow has been converted to kinematic

Water 2019, 11, 169 12 of 17 gully channels of different longitudinal gradients. The causes of the blockage were the reduction of flow area, the change of flow type and the decrease of flow velocity. The occurrence of the debris outburst, however, requires some necessary conditions, i.e., certain slope gradients on both sides and sufficient supply of solid materials. The slope in this area is sufficiently steep for the generation of large scaleWater surface2019, 1 x runoff,FOR PEER implying REVIEW a significant amount of gravitational energy. As such, the gravitational12 of 17 energy of abundant solid materials in the debris flow has been converted to kinematic energy, which canenergy, eventually which destroycan eventually the blockage. destroy The the potential blockage. energy The potential is not enough energy without is not water,enough and without solid materialwater, and would solid stop. material There iswould also the stop. contribution There is also of the the inertia contribution of the arriving of the solid-liquid inertia of the mixtures arriving at thesolid-liquid obstructed mixtures site. In some at the circumstances, obstructed site. the responsesIn some circumstances, of surface and undergroundthe responses water of surface were alsoand involvedunderground in the water outburst were phenomenon, also involved including in the ou thetburst seepage phenomenon, forces within including the blockage, the seepage the osmotic forces forcewithin of undergroundthe blockage, waterthe osmotic and the force erosion of byunderg surfaceround water. water However, and the in general,erosion by the surface blockage water. and outburstHowever, that in aregeneral, caused the by blockage natural topography and outburst have that little are influence caused by on natural human topography activities. have little influence on human activities.

(a) (b) (c)

FigureFigure 11. 11.Blockage Blockage types types of of debris debris ﬂows flows occurring occurring on on natural natural topography: topography: ( a(a)) constricted constricted section section of of gullygully channel,channel, (b(b)) arcarc gullygully channelchannel andand (c(c)) intersectionintersection betweenbetween thethe gullygully channelchannel ofof thethe largerlarger longitudinallongitudinal gradient gradient and and the the gully gully channel channel of of the the smaller smaller longitudinal longitudinal gradient. gradient.

WhenWhen a a debris debris flow flow spreads spreads into into human human settlements, settlements, the the blockage blockage followed followed by by outburst outburst that that is is causedcaused by by civil civil works works (such (such as as the the culvert culvert or or other other types types of of buildings) buildings) turns turns out out to to be be a a serious serious threat threat toto the the safety safety of of human human life life and and property. property. It It usually usually occurs occurs as as a a result result of of the the combination combination of of multiple multiple factors,factors, especially especially when when excessive excessive solid solid materials materials are are available available [ 55[55–57].–57]. If If it it is is a a normal normal surface surface runoff runoff oror a a small-scale small-scale debrisdebris flow,flow, as as shown shown in in Figure Figure 12 12a,a, it it would would pass pass through through the the culvert culvert without without the the blockageblockage because because of of the the lack lack of of sufficient sufficient solid solid material. material. As shownAs shown in Figure in Figure 12b, debris12b, debris flow offlow a large of a scalelarge cannot scale cannot be discharged be discharged through through the culvert the culver and at blockage and a blockage that acts that as aacts dam as temporarily a dam temporarily occurs. Duringoccurs. theDuring process, the process, as shown as in sh Figureown in 12 Figurec, some 12c, of some the debris of the flow debris runs flow across runs the across road the surface road bysurface overflow. by overflow. The blockage The aboveblockage the above road surface the road will surface eventually will beeventually destroyed be by destroyed the continuous by the comingcontinuous of the coming debris of flow the behind.debris flow One behind. example One of theexample outburst of the is depictedoutburst inis Figuredepicted 12 inc. FollowingFigure 12c. theFollowing outburst, the the outburst, release of the aggregated release of potential aggregated energy potential instantaneously energy instantaneously made the debris made flow the at adebris high velocity.flow at a In high this velocity. event, the In twothis dischargeevent, the diversionstwo discharge that diversions were caused that by were the blockagecaused by and the outburstblockage areand similar, outburst however are similar, they have however several they differences. have several The differences. discharge diversion The discharge occurring diversion at the rectangleoccurring culvertat the rectangle of the village culvert road of isthe because village theroad original is beca channeluse the original makes achannel turn right makes there. a turn The right discharge there. diversionThe discharge occurring diversion at the occurring arch culvert at the of thearch township culvert of road, the township however, road, is a special however, case is that a special is mainly case controlledthat is mainly by the controlled small transverse by the slopesmall gradient transverse of theslope road gradient surface. of Because the road some surface. residential Because quarters some areresidential in the downstream quarters are direction, in the downstream the debris direction, flow has causedthe debris severe flow damage has caused to the severe properties damage and to severalthe properties casualties, and which several could casualties, have been which avoided coul withd have a stronger been avoided discharge with capacity a stronger culvert. discharge capacity culvert.

Water 2019, 11, 169 13 of 17 Water 2019, 1 x FOR PEER REVIEW 13 of 17

(a)

(b)

(c)

Figure 12.12. DiagrammaticDiagrammatic sketches sketches of of different different sorts sorts of flow:of flow: (a) normal(a) normal surface surface runoff, runoff, mountain mountain flood orflood small-scale or small-scale debris flowdebris passing flow passing through thethrough culvert; the ( bculvert;) a large-scale (b) a large-scale debris flow debris passing flow through passing the culvertthrough and the ( cculvert) a special and circumstance (c) a special ofcircumstance a large-scale of debris a large-scale flow passing debris through flow passing the culvert through with the participationculvert with the of aparticipation great stone block. of a great stone block. 5. Conclusions 5. Conclusions Based on the field observations on the Puge event on 8 August 2017, the main conclusions can be Based on the field observations on the Puge event on 8 August 2017, the main conclusions can drawn as follows: be drawn as follows: (1) From the records in the meteorological station and Doppler weather radar analyses, the debris (1) From the records in the meteorological station and Doppler weather radar analyses, the debris flow occurred between 3:30 and 4:00 a.m. and the accumulative precipitation in the study area was flow occurred between 3:30 and 4:00 a.m. and the accumulative precipitation in the study area 49.2 mm on the day of the event. Because of the lack of rainfall measurements at high elevations, was 49.2 mm on the day of the event. Because of the lack of rainfall measurements at high it is deduced from the large size of the event and comparison with past debris flows that a large elevations, it is deduced from the large size of the event and comparison with past debris flows that a large amount of rainfall occurred in the middle and upper reaches of the Tongzilin gully.

Water 2019, 11, 169 14 of 17

amount of rainfall occurred in the middle and upper reaches of the Tongzilin gully. This event is a warning for establishing a complete and responsible rainfall monitoring system to guard against future potential geologic hazards. Prior to the occurrence of catastrophic consequences, it is possible to make a series of hazard prevention and mitigation works in advance to respond to the occurrence of debris flows. (2) According to estimates, the averages of peak velocity and peak discharge of debris flow are 6.9 m/s and 151 m3/s, respectively. It is quite a shocking motion compared to the recent events. The calculation results provide a design reference for the reconstruction of flood discharge structures. (3) Supply conditions, in addition to rainfall intensity and duration, are important factors that should be considered in predicting the occurrence of debris flows. In the current research, the ignoring of the supply conditions makes the prediction of the occurrence of debris flows meaningless. According to the consultation with local residents and the field investigation, large-scale surface runoff rarely occurs with the erosion effect before. Therefore, the supply condition of solid materials is abundant. For future government works, more prevention and control measures, such as gully channel dredging, gully bed reinforcing and slope stabilization, should be taken for channels with an abundant supply. (4) The transformation of natural topography by human activities is likely to increase the risk of geologic hazards. In the case of the Puge event, the blockage effect of two culverts made great contributions to the movement of debris flow by changing the flow direction. The construction of prevention and mitigation measures for geologic hazards must consider the occurrence and movement. Therefore, coordinating the relationship between humans and nature is the key to dealing with geologic hazards, including the rational layout of human living space, the thoughtful design of hazard prevention, mitigation construction and the effective conservation of soil and water.

Author Contributions: J.Z. and M.C carried out the field survey of this study and wrote the original manuscript. T.Z. and X.W. revised the original manuscript. X.L. and J.Z. conceived the original ideas and supervised the project. Funding: This research was funded by the National Key R&D Program of China (2017YFC1501102), the National Natural Science Foundation of China (51639007) and the Youth Science and Technology Fund of Sichuan Province (2016JQ0011). Acknowledgments: Critical comments by the anonymous reviewers greatly improved the initial manuscript. Conflicts of Interest: The authors declare no conflict of interest.

References

1. Chang, M.; Tang, C.; Van Asch, T.W.J.; Cai, F. Hazard assessment of debris flows in the Wenchuan earthquake-stricken area, South West China. Landslides 2017, 14, 1–10. [CrossRef] 2. Tang, C.; Zhu, J.; Li, W.L.; Liang, J.T. Rainfall-triggered debris flows following the Wenchuan earthquake. Bull. Eng. Geol. Environ. 2009, 68, 187–194. [CrossRef] 3. Zhou, J.W.; Cui, P.; Yang, X.G. Dynamic process analysis for the initiation and movement of the Donghekou landslide-debris flow triggered by the Wenchuan earthquake. J. Asian Earth Sci. 2013, 76, 70–84. [CrossRef] 4. Iverson, R.M. The physics of debris flows. Rev. Geophys. 1997, 35, 245–296. [CrossRef] 5. Iverson, R.M.; Reid, M.E.; Lahusen, R.G. Debris-flow mobilization from landslides. Annu. Rev. Earth Planet. Sci. 1997, 25, 85–136. [CrossRef] 6. Takahashi, T. Estimation of potential debris flows and their hazardous zones: Soft countermeasures for a disaster. J. Nat. Disaster Sci. 1981, 3, 57–89. 7. Coe, J.A.; Kinner, D.A.; Godt, J.W. Initiation conditions for debris flows generated by runoff at Chalk Cliffs, central Colorado. Geomorphology 2008, 96, 270–297. [CrossRef] 8. Tognacca, C.; Bezzola, G.R.; Minor, H.E. Threshold criterion for debris-flow initiation due to channel-bed failure. In Proceedings of the 2nd International Conference on Debris-Flow Hazard Mitigation, Taipei, Taiwan, 16–18 August 2000; pp. 89–97. Water 2019, 11, 169 15 of 17

9. Gregoretti, C. Experimental evidence from the triggering of debris flow along a granular slope. Phys. Chem. Earth Part B Hydrol. Ocean. Atmos. 2000, 25, 387–390. [CrossRef] 10. Gregoretti, C. The initiation of debris flow at high slopes: Experimental results. J. Hydraul. Res. 2000, 38, 83–88. [CrossRef] 11. Hu, W.; Dong, X.J.; Wang, G.H.; van Asch, T.W.J.; Hicher, P.Y. Initiation processes for run-off generated debris flows in the Wenchuan earthquake area of China. Geomorphology 2016, 253, 468–477. [CrossRef] 12. Berti, M.; Simoni, A. Experimental evidences and numerical modelling of debris flow initiated by channel runoff. Landslide 2005, 2, 171–182. [CrossRef] 13. Imaizumi, F.; Sidle, R.C.; Tsuchiya, S.; Ohsaka, O. Hydrogeomorphic processes in a steep debris flow initiation zone. Geophys. Res. Lett. 2006, 33, 229–237. [CrossRef] 14. Cannon, S.; Gartner, J.E.; Wilson, R.C.; Bowers, J.C.; Laber, J.L. Storm rainfall conditions for floods and debris flows from recently burned areas in Southwestern Colorado and Southern California. Geomorphology 2008, 96, 250–269. [CrossRef] 15. Gregoretti, C.; Dalla Fontana, G. The triggering of debris flow due to channel-bed failure debris flow in some alpine headwater basins of the Dolomites: Analyses of critical runoff. Hydrol. Process. 2008, 22, 2248–2263. [CrossRef] 16. Theule, J.I.; Liebault, F.; Loye, A.; Laigle, D.; Jaboyedoff, M. Sediment budget monitoring of debris flow and bedload transport in the Manival Torrent, SE France. Nat. Hazard Earth Syst. Sci. 2012, 12, 371–749. [CrossRef] 17. McCoy, S.W.; Kean, J.W.; Coe, J.A.; Tucker, G.E.; Staley, D.M.; Wasklewicz, W.A. Sediment entrainment by debris flows: In situ measuremeants from the head-waterscof of a steep catchment. J. Geophys. Res. Earth Surf. 2012, 117, F03016. [CrossRef] 18. Kean, J.W.; McCoy, S.W.; Tucker, G.E.; Staley, D.M.; Coe, J.A. Runoff-generated debris flows: Observations and modeling of surge initiation, magnitude, and frequency. J. Geophys. Res. Earth Surf. 2013, 118, 2190–2207. [CrossRef] 19. Hurlimann, M.; Abanco, C.; Moya, J.; Vilajosana, I. Results and experiences gathered at the Rebaixader debris-flow monitoring site, Central Pyrenees, Spain. Landslides 2014, 11, 939–953. [CrossRef] 20. Tiranti, D.; Deangeli, C. Modeling of debris flow depositional patterns according to the catchments and sediment source areas characteristics. Front. Earth Sci. 2015, 3, 8. [CrossRef] 21. Ma, C.; Deng, J.; Wang, R. Analysis of the triggering conditions and erosion of a runoff-triggered debris flow in Miyun County, Beijing, China. Landslides 2018, 15, 2475–2485. [CrossRef] 22. Qi, S.; Vanapalli, S.K. Hydro-mechanical coupling effect on surficial layer stability of unsaturated expansive soil slopes. Comput. Geotech. 2015, 70, 68–82. [CrossRef] 23. Qi, S.; Vanapalli, S.K. Influence of swelling behavior on the stability of an infinite unsaturated expansive soil slope. Comput. Geotech. 2016, 76, 154–169. [CrossRef] 24. Qi, S. Numerical Investigation for Slope Stability of Expansive Soils and Large Strain Consolidation of Soft Soils. Ph.D. Thesis, University of Ottawa, Ottawa, ON, Canada, 2017. 25. Theule, J.I.; Liebault, F.; Laigle, D.; Loye, A.; Jaboyedoff, M. Channel scour and fill by debris flows and bedload transport. Geomorphology 2015, 243, 92–105. [CrossRef] 26. Reid, M.E.; Coe, J.A.; Dianne, L.B. Forecasting inundation from debris flows that grows volumetrically during travel, with application to the Oregon Coast Range, USA. Geomorphology 2016, 273, 396–411. [CrossRef] 27. Gregoretti, C.; Degetto, M.; Bernard, M.; Boreggio, M. The debris flow occurred at Ru Secco Creek, Venetian Dolomites, on 4 August 2015: Analysis of the phenomenon, its characteristics and reproduction by models. Front. Earth Sci. 2018, 6, 80. [CrossRef] 28. Gregoretti, C.; Stancanelli, L.M.; Bernard, M.; Boreggio, M.; Degetto, M.; Lanzoni, S. Relevance of erosion processes when modelling in-channel gravel debris flows for efficient hazard assessment. J. Hydrol. 2019, 569, 575–591. [CrossRef] 29. Gregoretti, C.; Degetto, M.; Bernard, M.; Crucil, G.; Pimazzoni, A.; De Vido, G.; Lanzoni, S. Runoff of small rocky headwater catchments: Field observations and hydrological modeling. Water Resour. Res. 2016, 52, 8138–8158. [CrossRef] 30. Imaizumi, F.; Hayakawa, Y.S.; Hotta, N.; Tsunetaka, H.; Ohsaka, O.; Tsuchiya, S. Relationship between the accumulation of sediment storage and debris-flow characteristics in a debris-flow initiation zone, Ohya landslide body, Japan. Nat. Hazards Earth Syst. Sci. 2017, 17, 1923–1938. [CrossRef] Water 2019, 11, 169 16 of 17

31. Iverson, R.M.; Schilling, S.P.; Vallance, J.W. Objective delineation of lahar-hazard zones downstream from volcanoes. Geol. Soc. Am. Bull. 1998, 110, 972–984. [CrossRef] 32. Rickenmann, D. Empirical relationships for debris flows. Nat. Hazards 1999, 19, 47–77. [CrossRef] 33. Rickenmann, D.; Weber, D. Flow resistance of natural and experimental debris flows in torrent channels. In Proceedings of the Second International Conference on Debris Flow Hazard and Mitigation, Taipei, Taiwan, 16–18 August 2000; pp. 245–254. 34. Santi, P.M.; de Wolfe, V.G.; Higgins, J.D.; Cannon, S.H.; Gartner, J.E. Sources of debris flow material in burned areas. Geomorphology 2008, 96, 310–321. [CrossRef] 35. Flores, J.; D’Alpaos, A.; Squarzoni, C.; Genevois, R.; Marani, M. Recent changes in rainfall characteristics and GIS and their influence on threshold for debris flow triggering on Dolomitc area of Cortina d’Ampezzo, North-eastern Italian Alps. Nat. Hazard Earth Syst. Sci. 2010, 10, 571–580. [CrossRef] 36. Chiarle, M.; Iannotti, S.; Mortara, G.; Deline, P. Recent debris flow occurrences associated with glaciers in the Alps. Glob. Planet. Chang. 2007, 56, 123–136. [CrossRef] 37. Ren, D. The devastating Zhouqu storm-triggered debris flow of August 2010: Likely causes and possible trends in a future warming climate. J. Geophys. Res. Atmos. 2014, 119, 3643–3662. [CrossRef] 38. Tang, C.; Rengers, N.; van Asch, T.W.J.; Yang, Y.H. Triggering conditions and depositional characteristics of a disastrous debris flow event in Zhouqu City, Gansu Province, Northwestern China. Nat. Hazards Earth Syst. Sci. 2011, 11, 2903–2912. [CrossRef] 39. Tang, C.; Zhu, J.; Ding, J.; Cui, X.F.; Chen, L.; Zhang, J.S. Catastrophic debris flows triggered by a 14 August 2010 rainfall at the epicenter of the Wenchuan earthquake. Landslides 2011, 8, 485–497. [CrossRef] 40. Yu, B.; Ma, Y.; Wu, Y. Case study of a giant debris flow in the Wenjia gully, Sichuan Province, China. Nat. Hazards 2013, 65, 835–849. [CrossRef] 41. Zhou, J.W.; Cui, P.; Yang, X.G.; Su, Z.M. Debris flows introduced in landslide deposits under rainfall conditions: The case of Wenjiagou gully. J. Mt. Sci. 2013, 10, 249–260. [CrossRef] 42. Xie, H.; Zhong, D.L.; Jiao, Z.; Zhang, J.S. Debris Flow in Wenchuan Quake-hit Area in 2008. J. Mt. Sci. 2009, 27, 501–509, (Chinese Reference). 43. Liao, Y.F.; Li, Y.; Chen, X.Y.; Feng, J.N. Study and application of glaze environmental influence factors—Taking hunan as an example. J. Catastrophol. 2011, 26, 76–81, (Chinese Reference). 44. Cui, P.; Lin, Y. Debris-flow treatment: The integration of botanical and geotechnical methods. J. Resour. Ecol. 2013, 4, 97–104. 45. Gan, B.R.; Liu, X.N.; Yang, X.G.; Wang, X.K.; Zhou, J.W. The impact of human activities on the occurrence of mountain flood hazards: Lessons from the 17 August 2015 flash flood/debris flow event in Xuyong County, south-western China. Geomat. Nat. Hazards Risk 2018, 9, 816–840. [CrossRef] 46. Chen, N.; Chen, M.; Li, J.; He, N.; Deng, M.; Iqbal Tanoli, J.; Cai, M. Effects of human activity on erosion, sedimentation and debris flow activity—A case study of the Qionghai Lake watershed, southeastern Tibetan Plateau, China. Holocene 2015, 25, 973–988. [CrossRef] 47. Wang, G.L. Lessons learned from protective measures associated with the 2010 Zhouqu debris flow disaster in China. Nat Hazards 2013, 69, 1835–1847. [CrossRef] 48. Bovis, M.J.; Jakob, M. The role of debris supply conditions in predicting debris flow activity. Earth Surf. Process. Landf. 1999, 24, 1039–1054. [CrossRef] 49. Berger, C.; Mcardell, B.W.; Schlunegger, F. Direct measurement of channel erosion by debris flows, Illgraben, Switzerland. J. Geophys. Res. Earth Surf. 2011, 116, 93–104. [CrossRef] 50. Kean, J.W.; Coe, J.A.; Coviello, V.; Smith, J.B.; Mccoy, S.W.; Arattano, M. Estimating rates of debris flow entrainment from ground vibrations. Geophys. Res. Lett. 2015, 42, 6365–6372. [CrossRef] 51. Prochaska, A.B.; Santi, P.M.; Higgins, J.D.; Cannon, S.H. A study of methods to estimate debris flow velocity. Landslides 2008, 5, 431–444. [CrossRef] 52. Hungr, O.; Morgan, G.C.; Kellerhals, R. Quantitative analysis of debris torrent hazards for design of remedial measures. Can. Geotech. J. 1984, 21, 663–677. [CrossRef] 53. Lanzoni, S.; Gregoretti, C.; Stancanelli, L.M. Coarse-grained debris flow dynamics on erodible beds. J. Geophys. Res. 2017, 122, 592–614. [CrossRef] 54. Stancanelli, L.M.; Musumeci, R.E. Geometrical Characterization of Sediment Deposits at the Confluence of Mountain Streams. Water 2018, 10, 401. [CrossRef] Water 2019, 11, 169 17 of 17

55. Luo, P.; Mu, D.; Xue, H.; Ngo-Duc, T.; Dang-Dinh, K.; Takara, K.; Nover, D.; Schladow, G. Flood inundation assessment for the Hanoi Central Area, Vietnam under historical and extreme rainfall conditions. Sci. Rep. 2018, 8, 12623. [CrossRef][PubMed] 56. Luo, P.; Zhou, M.; Deng, H.; Lyu, J.; Cao, W.; Takara, K.; Nover, D.; Schladow, G. Impact of forest maintenance on water shortages: Hydrologic modeling and effects of climate change. Sci. Total Environ. 2018, 615, 1355–1363. [CrossRef][PubMed] 57. Zhou, J.W.; Yang, X.G.; Hou, T.X. An analysis of the supply process of loose materials to mountainous rivers and gullies as a result of dry debris avalanches. Environ. Earth Sci. 2017, 76, 452. [CrossRef]

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).