#76 Modelling with HEC-HMS

Copy Help
  • Public/Private: Change the visibility of this video on your My Videos tab
  • Save/Unsave: Save/Unsave this video to/from your Saved Videos tab
  • Copy: Copy this video link to your system clipboard
  • Email: Copy this video link to your default email application
  • Remove: Remove this video from your My Videos or Saved Videos tab
Watch at: 00:00 / 00:00:20subscribe to this youtube channel byclicking the subscription button belowwelcome to this australian water schoolwebinar we're so glad you could join ustoday to discussmodeling with heck hms presented bythomas brower michael bartles craigpricestephen joins this is the hydrologicmodeling system designed to simulateWatch at: 00:20 / 00:40complete hydrologic processeswell let's get on with this my name'strevor filler i'm the nationalpartnership manager here at ice warmand the webinar chair but it's beenreally good to see the response of thiswebinar today you can see on the screensthereright today four presenters uhthomas brower michael bartles craigprice and stephen joyneswe're so glad you could join usWatch at: 00:40 / 01:00gentlemen um you can seea fantastic amount of experience herewelcome to you all i want to ask youjust one question each before wegot this off what drives you to thiswhat motivates you to umto get right into this hms maybe we'llstart with um ukraine crywell it may be no surprise that i'm moreWatch at: 01:00 / 01:20of a hydraulics guy than a hydrology guybut uh steve and i have been uhteam teaching a hms and hec raz courseonce a year in new zealand um which hasbeen a face-to-face courseand now we're venturing out doing thison our own uh with with ice warm now asa partner to do itthrough the online system so we'repretty excited about that um it's beenvery goodWatch at: 01:20 / 01:40to be able to do this um with uh youknow and and to get involved in thisbut really frankly um i won't uh try topull the wool over anybody's eyesuh frankly i i'm i use uh hms forboundary conditions sothat's what i use it for so um i'll passalong to everybody else i need the uhhydrographs to put into my hecarasmodels umand uh but i am excited uh because um toWatch at: 01:40 / 02:00we we use thosegrass models um as rainfall runoffmodels as well and i am really excitedto find out what's new in the upcomingversion 4.4 soi don't want to take much more time thanthat pass it off to the other guysyou sound like you're you should havethe background that i've gotin my screen but you're like you're notinthe us yes i was just i was just thereWatch at: 02:00 / 02:20actually in san francisco a couplemonths ago actually at the hecoffices and um you know i've been inaustralia about 12 years nowand um but i do get back to the usoccasionally that's uhi went to berkeley for my uh schoolingso just across the bay i think you mightactually see it on the left there yesnot far from here yeah right yeah yeahWatch at: 02:20 / 02:40steve yourself from the beautiful newzealandyeah hi everybody this is steve jonesherebeen teaching akhms for at least 10years now i've been using software for20 yearsat least out of necessity but over timeit's really interesting to use it's sosimpleand to be honest i think i practicallyWatch at: 02:40 / 03:00use it every single day of my workinglifeso i'd like to think and know it insideout butyou never know everything about softwareso it's going to be an interestingthing for me but um always happy toanswer questions as you need itthank you very much for joining us todayand uh craig and steve will be on theuh q a chat lines right through thisWatch at: 03:00 / 03:20whole hour sodo write your questions they'll beanswering them they'll be talking withyou about thatright up to you michael what what bringsyou to heck hms andthis whole area of hydrologic modellingso um i really enjoy the big problemsright so i enjoyworking on the models that are hundredsof thousands of square miles orWatch at: 03:20 / 03:40should i say square kilometers we don'tsee what the mattress is very often herein the usbut i like the very very large problemsum it's always been attracted to themand that is mostly in the realm ofhydrologic modeling though i do dabblein hydraulic modeling quite a bitas well yeahyeah i guess for me i once again likeWatch at: 03:40 / 04:00mike said i was always attracted to thebig problems and the army corps isusually doing the big problems and youcome to a place like hecand you're doing the hardest of the bigproblemsand so that's all always been fun for meand i'll saythe venn diagram between computerscience and hydrology is kind of theworld i operateWatch at: 04:00 / 04:20in and that's a really fun world for meit was a combinationof my interests yep well uh that'sthis is a good start we'll get we'll getstuck into this now the way we're goingto do this iscrow's going to pick up going to kickoff with a couple of interestslides and and just talk about theoverall picture then we're going to handover to tomuh and michael and then we'll have timeWatch at: 04:20 / 04:40for q ac will also be on boardactively working with the q a linebut my background today is reallydedicated to you two gentlemen becauseyou're not far from here i understandyou actually live quite close to thisthat's trueuh we work out of the hydrologicWatch at: 04:40 / 05:00engineering center davis california sowe're about two hours eastof where this picture is taken one of myfavorite places in the world actually iswhere you'resupposedly standing finallyno adelaide south australia not quitethe same but pretty good place to livethough i can tell youokay no more of this let's get stuckinto this uh right over did you craveWatch at: 05:00 / 05:20missed anything let me know otherwise uhyou you lead us away craigexcellent um first thing is always ani.t check can you see my screen with theuhuh presenters fantastic excellentokay and i want to keep my remarks asshort as we possibly can because we wantto dive right into tom and michael's uhcomments um what i wanted to do thoughWatch at: 05:20 / 05:40is just uh kind of introduce where we'reat with the hydrologic engineeringcenter we have had probablysix or eight of these webinars um on uhhec ras so on the hydrologic engineeringcenter the hecand the purist will say hec instead ofheck um we alluh at least on this webinar there's alot of people who have sat in onWatch at: 05:40 / 06:00hecraz um discussions um we've had overfifty thousand views of our uh hecarim'swebinarsas of tomorrow when this one is postedwe'll probably have our firsthms view on youtube but i wanted to justumset the stage for what we're looking athere um hec-ras came from uh the hec2-pack it's a fortran package whichWatch at: 06:00 / 06:20computedwater surface profiles and some of youmay be my age or older andremember that but there's also heckthree heck four heck five heck sixum if you want to learn a little moreabout those um we cani'll show you a link here in just aminute but um what was heck one what wasthe predecessor to this heck radwe can answer chris question i see whereWatch at: 06:20 / 06:40craig was going with this sothe predecessor to hms was actually hec1so if anybody's used hec1then hec hms a lot of it will make senseto themand if not we'll clarify what hms isto the world today so aswas introduced earlier my name is tom myWatch at: 06:40 / 07:00name is mike bartlesand we are hydraulic engineers from thehydrologic engineering centeras we said in davis california so hec isa partof the army corps of engineers withinthe united statesunderneath the department of defense sotoday thank you all first and foremostfor joining on certainly appreciate theopportunity to talk with you about htcWatch at: 07:00 / 07:20hmsand without further ado let's get intoit so here's an overview of ourpresentationtoday we're going to talk about hms alittle bit of an overview and itsvarious capabilitieswe're also going to talk quickly aboutthe differences between hydrologic andhydraulic modelingsince that was brought up earlier thenthrough demonstration we're going toshow how an hms model can be createdWatch at: 07:20 / 07:40parameterized calibrated and calibratedusing observed data quicklyand then finally we're going to closewith a quick discussion of enhancementsthat we plan on including within hms inthe futureand then so do some q a afterwardsokay first let's start with hydrology sohydrology is the study of water or moreWatch at: 07:40 / 08:00formallythe scientific study of the movementdistribution and quality of waterwe have a nice picture here that i thinkcaptures hydrology well we seeprecipitation falling from some cloudswe see a land surface that has apotential for infiltrationwe see some vegetation with the capacityWatch at: 08:00 / 08:20to transpire waterand we see a lake a river system herethat it probably flows downstream to theocean at some pointso all of this is captured in thispicture and this is hydrologynow let's take the same thing we justlooked at in an image form and let'slook at it in more of a chartWatch at: 08:20 / 08:40so we in this chart we see somehydrologic processeswell all of these processes are whatwe're modelingin hms so we think see things likeprecipitation we have methods to modelprecipitation we see surface runoff thisisthe major thing we're capturing in hmswe have lost rates thatWatch at: 08:40 / 09:00subtract from the precipitation excesswe have things like lakes and reservoirsand hms riverseven things like evapotranspiration andevaporationso a few bullet points on hmsit's applicable for most regions of theworld it's the way it's designed it canWatch at: 09:00 / 09:20beit can and has been applied all aroundthe worldit's both we have and so for eachprocess i guess one point of distinctionthat i often come tois people sometimes will say oh hms andcall it a modeland the point of distinction i like tomake is that hms is actually a programWatch at: 09:20 / 09:40that you do hydrologic modeling in sowith each hydrologicprocess we have multiple approaches formodeling that processand so the end model the the resultmodel in hms is actuallyit's nearly always a custom model thatthe user has builtso we'll talk about things like we haveempirical methods and physically basedWatch at: 09:40 / 10:00methodsfor for each hydrologic process peoplewill talkwhen they talk modeling they tell is itevent or continuous wellhms can do either depending on how whichmodeling methods you select and how youparameterize itand then one another that comes up is itgridded or lumpedthose terms fly around in the modelingworld andWatch at: 10:00 / 10:20hms can do either of those we havegridded methodswe have lumped parameter i wouldsometimes you can you could call thatuh semi-distributed so hms we haveapproaches for doing both of thosesome of the strong suits of hms is datahandlingso temporal aggregation anddisaggregationhms handles all that automatically so ifWatch at: 10:20 / 10:40you havedaily precipitation data and run yourmodel at a six hour time stephms handles that another thing that hmsdoes well is handling unit systems soin the states we often work in u.scustomary units around the world it'soften metricand so hms handles theWatch at: 10:40 / 11:00conversions between those unit systemswithout automatically without any inputfrom the useruh and so another beneficial thing abouthms is we havesome compute types that are orientedtoward hydrologic studiesso the most common of which is asimulation run we also have anWatch at: 11:00 / 11:20optimization trial where you're able tooptimizeparameters we have a forecastalternative that's set up forforecastingwe do depth area analyses which aretypically paired with afrequency storm for designing hydrologicand hydraulic structuresand then we do an uncertainty analysisWatch at: 11:20 / 11:40where we can start to consider thingslike parameter uncertaintyso to have a quick discussion about thedifferences between hydrologic andhydraulic modelingum so in the past htc one hec2 worlduh there were very well definedboundaries between hydrologic modelsnow within the hms and htc raspberryworld that we live in today thoseWatch at: 11:40 / 12:00boundaries have started to become alittle bit moreblurred for instance hydraulic modelshave started to allow for the simulationof rainfall on the overland plain as youguys have seen in various otherpresentationsalso hydrologic models have started tobecome capable of including hydraulicroutingbut before choosing a hydrologic modelor a hydraulic model for use in studyit's important to take into account theWatch at: 12:00 / 12:20considerations of the studyitself so that includes things like runtimesso for instance hydrologic models tendto require less time to complete asimulation due to their inherentsimplifications whereashydraulic models take a bit more timebecause they're inherently morecomplicatedthe nature of the system you should alsotake that into accountfor instance very very flat channelWatch at: 12:20 / 12:40slopes so in the uswe usually apply a rule of thumb abouttwo feet per mile anything less thanthathydraulic routing tends to be the bestchoice that's usually the best choiceanyways but hydrologic routingreally starts to fall apart and doesn'tdo a great job of capturingtrue attenuation and translation effectsof flood waves through systemsso if you're dealing with systems likeWatch at: 12:40 / 13:00that you might want to be looking moretowards the hydraulic modeling realmand also level of detail and desiredoutputs so if you're interested ininundation extentsor arrival times hydraulic models aregoing to be your best betsto determine those those outputsas a quick example here's an exampleof hydrologic and hydraulic models beingWatch at: 13:00 / 13:20used in tandem with one anotherso this is a screen shot of the uppersusquehanna river watershedin new york and pennsylvania in the u.swithin the core water management systemthat we use within the army corps ofengineers to manage our water managementprojectslike i said manage a whole bunch oftimes in there but either way so in thiswatershed we have two usace owned andWatch at: 13:20 / 13:40operated projectsthat are in working in tandem to reduceflooding risk to populated areasso we run hydrologic simulations on adaily basis to forecast inflowhydrographs to both these damsas well as hydrographs for downstreamareas this is done within our hydrologicmodel due to time constraintsdesired level of accuracy and desiredoutputs that we needWatch at: 13:40 / 14:00but hydraulic simulations are computedonly when necessaryusually when stages approach or exceedsome defined flood threshold downstreamand populated areas this is done becausethe hydrologic or hydraulic simulationstake longerto complete and inundation extentsthey're not always requiredduring normal operations so we handoutput from one to the otherusing our data storage system hc dssWatch at: 14:00 / 14:20from one model to the other rightall right at this point we're going toleave the powerpoint presentation we'reactually going to do a little bit ofdemonstrationso what i'm going to do is i'm going toopen htc hmsnow i'm going to begin a brand newproject so i want to placeWatch at: 14:20 / 14:40this project in this locationand i want to call it punxsutawneyso if anybody has ever seen the awesomemoviegroundhog day you'll understand wherethis place is if you don't definitelylook it upit's a great movie to watch so now thati have a brand new project within hccWatch at: 14:40 / 15:00hmsi want to create and import a terraindataso i'm going to give it a name let'scall it nad 10 meternad stands for national elevation dataset and 10 meters is the horizontalresolutionso this data actually covers theentirety of the continental unitedstatesi want to getWatch at: 15:00 / 15:20this data set so it's a geotiff rasternow that i've imported thati can look and see that i have thisterrain file that's been imported and ican now create a basic modelso i'm going to call this base and modelpunks just a shorthand name[Music]Watch at: 15:20 / 15:40and i want to associate the previouslyadjusted terrain datai'm going to select it i'm going to sayskip because i just want to transfer thecoordinate systemfrom the terrain data now that i've donethat you can see my terrain datain the background and i can begin doingall sorts of things one of those thingsWatch at: 15:40 / 16:00being i want topre-process syncs so what this is doingis creating a hydrologically correcteddemso one that doesn't include artificialpits you'll see that i've created twonew layersone is a location of all the sinkswithin the watershed that have beenfilledand also a resultant dem that has beenWatch at: 16:00 / 16:19created over thatnext step i want to pre-process thedrainage what's happening in thisprocessas i'm computing flow directions andflow accumulationsso if you're used to how geo hmsand or archive used to work you knowthat those are both created as part ofthat process as wellso let me turn off the flow accumulationWatch at: 16:19 / 16:40raster and show you what the flowdirection raster looks like so this isjust computingwhich way water were to flow and theflow accumulationtakes into account everything upstreamof that to show you the drainage networkso you can start to seedendritic channels and where water wouldactuallyflow if it were flowing on the overlandplain so i'm going to turn that one offi'm going to go back to my base terrainWatch at: 16:40 / 17:00and nowi want to identify streams so this isasking for a threshold to say this is astream or notand i'm going to input 70 squarekilometersand now it's going to determine whereenough drainage area is accumulated thatexceeds 70 square kilometers and i'mgoing to call that a streamand now i want to add a break point andWatch at: 17:00 / 17:20i'm going to go to a location where astream gage exists so at this locationthe united states geologic survey iscapturingessentially stream flow we'll just callit thatat every single time step so 15 minutetime step in this caseor hourly depending upon your specificWatch at: 17:20 / 17:40location now that i have a breakpointidentifiedi can say delineate me these elementsand in very very short order i will havecreateda basin model that contains sub basinsand routing reaches so all these guysare sub basins these are your catchmentswhere the precipitation runoff processWatch at: 17:40 / 18:00is modeledand these are streamflow routing reacheswhere channel routing processes aresimulatednow that i have this stuff i can startto get specific for instancethis tributary right here is actuallycalled stump creek so i'm going torename that sub basini'm going to call it steep creek thisone is also calledeast branch mahoney creek so i'm goingto simplify that i'm going to call itWatch at: 18:00 / 18:20east branch mahoning creekand i also want to merge these sidestogether so i just want to make onesingle sub basin here so i select thosetwoi'm going to merge those elements i'mgoing to select those twoi want to merge those elements and now ican rename this guythis is actually my homing creek[Applause]Watch at: 18:20 / 18:40and i want to make another junctionright here and i'm going to call thisconfluence and i want to connectthese sub basins to that junction so ican rename orre-select their downstream linkage andnow that they're selectedi can also select downstream here forreach toand finally i want to select these twoWatch at: 18:40 / 19:00routing reach elementsand i'm going to merge them together ishould say one more thing i want torename this guy and i want to call thatsyncgauge and one more thing i lied againi'm going to call this money creek andi'm going to give it an r1 designationto mean reachso in one fell swoop well oneWatch at: 19:00 / 19:20a couple swoops i have created acomplete basin modelum discretized everything that goes onin here one last stepi want to create a grid region and iwant to call thisshg which stands for standard hydrologicgrid which is the option hereyou'll see what happens with this one inthe next stepWatch at: 19:20 / 19:40i'm going to click finish and i'm goingto set that as the defaultgrid region all rightokay get over tommy all right so nowthat mike's created a basin model i'mgoing to prepare a meteorologic boundarycondition so we can dump some precip onthis basin and we're also going to applysome temperatures so we're going to do ashort continuous simulationWatch at: 19:40 / 20:00to import gridded precipitation data wehave thisvortex import utility so this is astandalone set of small utilities it'skind of a test bed for features thatwill probably incorporate into hmsin the future but for now we have thisWatch at: 20:00 / 20:20little wizard and you'll see it's prettyeasy to usei'm going to select these grid files themetdata importerall it does is it reads common griddeddata formatsone of which is grid netcdf is anotherhdf5 is another you might be familiarwithso we'll just select all of these filesWatch at: 20:20 / 20:40and now the program is just going toloop over the files and take a lookinside and see what variable is in thereso right now i'm importing mrms datait's a prettygood data set for north america ifyou're not in north america you mightlook out to some other data sources seewhatwhat is available in your neck of theWatch at: 20:40 / 21:00woods another one that you mightconsider using is nasa's gpm dataso i'm going to select that variable andnow i have the option to clipthese grids to my basin and so righthere i just have a shapefile that is mybase and extenti'm also going to set the projection uhthe target projection for the grids iWatch at: 21:00 / 21:20want these to bean shg if you're working in hms i wouldstrongly suggest using utm or shg ifyou're working in statesbut you can use any projection you wanti'm justsuggesting these and then i'm going toselect a target cell sizeWatch at: 21:20 / 21:40so that's going to be i'm going to use acell size of 2 000.and now we just need to say where do iwant to write these grids we're going towrite these to hec dss which is thedata storage format that all hcc toolsuseso i have an empty dss file i'm justgonnatell the grids to right there so theseWatch at: 21:40 / 22:00grids are gonnait's quite a bit of data to churnthrough so each grid it's going to gothroughread the data it's going to clip itreproject itresample it and then write it so thiswill take a minutei've already created some grids so we'lljust move forward with the grids thati've created and then we'll come backaround at the end and take a look atsome of the dataso what i'm going to do now is addWatch at: 22:00 / 22:20gridded data so i'm going to add aprecipitation grid seti'm going to name this qpebecause that's the data we're usingnow i'm going to link the dss filethat's associated with thisWatch at: 22:20 / 22:40so i've linked the dss file now i justneed to link a dss pathyou can select any but i usually go withthe first so i've linked the precip dataWatch at: 22:40 / 23:00now i'm going to create a temperaturegridi'm going to name this rtma that's thetemperature data that i'm usingnow i'm going to link the dss fileand i'm going to select a recordWatch at: 23:00 / 23:20okay now that i've added my data i'mgoing to create a meteorologic modelso to do that i do components met modelmanagerand create a new met model i'm going toname thisseptember 2018.Watch at: 23:20 / 23:40now on this model i have the opportunityto select the differentmethods that i want to use so one thingi'm going to do is i'm going to usegridded presetanother thing i'm going to do is usegrid hammond evapotranspirationnow on this gridded precipitation methodWatch at: 23:40 / 24:00i can select my precipitation gridsand in this evapotranspiration methodi can use rtma temperature grids anotherthing that i'm going to do with this metmodel is link it to the basin model thati'm usingso i'm just going to tell it to use thesubbasins from the punxsutawney basinmodelWatch at: 24:00 / 24:20okay i think i have one last thing andthat's the linka gauge here so i'm going to bring instreamflow data that's observedand to do that i'm going to create atime series gagethis is a discharge gauge and i'm goingto name thisusgs funks for punxsutawneyWatch at: 24:20 / 24:40so i've added a time series gaugenow i'm going to link this to theobserve flow recordin dsswhich looks something like that now iWatch at: 24:40 / 25:00just want to link this synci want to set this observed gaugeso now when we run a simulation we'llhave observed flow to compare tookay so back over to vortex we see thatour import has completedso i'll just close the wizard and nowlet'sWatch at: 25:00 / 25:20take a look inside ofdss so here's all our records importedinto dssand if i go to the 9th of september wasthe most active period so let's justselect a handful of records and then wecan graph themWatch at: 25:20 / 25:40and we start to see that it looks likewe have for thishour we have point four millimeters inthat particular celland i'll just step through a couple soyou can see what this looks like so thisiswe've created a precipitation boundarycondition for our basin so that's whatthe data looks like inside of gssWatch at: 25:40 / 26:00right and now i'm going to quicklyparameterize this modelso first things first i want to set myprocesses for eachsub basin and routing reach element iwant to usedeficit and constant losses i want toWatch at: 26:00 / 26:20give parameter values to each sub basinof 50 millimetersof initial deficit maximum storage of200 millimeters and a constant loss rateof 2.5 millimeters per houri also want to set the canopy methodto use the simple canopy methodand parameterize that to use an initialWatch at: 26:20 / 26:40storage of 0.1 percent or zero percentand a max storage of 0.1 millimeters soit's essentially nothingand simple uptake so what this will dois allow me to actually extract waterfrom the soil for transformi want to set this to use the modifiedclark methodWatch at: 26:40 / 27:00and give that parameter 6 7 8 910 values for each individual sub basinso time of concentrationand watershed storage coefficientfor base flow i want to use the linearreservoir methodso this guarantees conservation of massand i want to use two layers for eachWatch at: 27:00 / 27:20sub basin set the discharge the initialtype to discharge per areainitially have zero coming out of thefirst layer and .015cubic meters per second per squarekilometer out of the second layer50 of infiltrated water should go to thegroundwater one layerWatch at: 27:20 / 27:40and use a coefficient a routingcoefficient of 20 hours50 goes to groundwater 2 with a storagecoefficient of 200 hoursthat's all i've got to do for that onemore thing i want to set my routingreach to use the musking and routingmethodand i want to have a k value of 0.05 anda mustangx of 0.25 and a number of subregionsWatch at: 27:40 / 28:00of 5. now i want to create a controlspecificationsfor the simulation at hand i'm going tocall that september 2018.now i specifythat i want to start on the 1st ofWatch at: 28:00 / 28:20september2018 zero zero zero zero and i want toend on the 28th of september2018 zero zero zero as well and a timestep of one houri'm going to set that to be acomputation pointand now i can run i can create asimulation runi'm going to call this september 2018Watch at: 28:20 / 28:40i need to select a base and model i onlyhave one i need to select a met model ionly have one and i need to select acontrol specification i only have one sothose choices are easyand now i can make a simulation runfirst time i runsince i'm using the modified clarkmethod it needs to computetime area histograms essentially andWatch at: 28:40 / 29:00travel time indices for each individualsub basin so that's what this is doingthis will only happen one timeso it needs to do this one time and thennever againunless i make modifications to the basinmodel itself you'll seesubsequent simulations go much much muchfasterso now that simulation is complete i'llkick it back over to tommyokay so we've ran our first simulationWatch at: 29:00 / 29:20and i think we're about10 15 minutes into the demo so that'spretty quick let's take a look at someof the results soone of the opportunities you have toview results is straight out of thisbasin map so i just right clicked on thesync element we have here and i have theopportunity to viewgraphical results so here we see ourWatch at: 29:20 / 29:40graphical results not looking bad forour first runanother type of result we have is atable so i can go in here and i can viewa summary tableand i see things like a peak dischargea volume i can change the units on thisi also see things like an ashcliff efficiency so 0.95 hey we can justWatch at: 29:40 / 30:00stop now that's pretty gooduh percent bias so that gives anindication of volume so we have someperformance metrics and these standardsummary tablesanother way i can access the results isfrom the resultstab so i've just switched from thecomponents tab which isan outline of the watershed and now iWatch at: 30:00 / 30:20have results here on this tab so i canjust start to click throughelement nodes and view different typesof results so here's just the resultsforeast branch mahoney creek and if iscroll down we can go back to our gaugeand access the same results that we justsaw from the basin maplet's keep this here for a second and doa briefWatch at: 30:20 / 30:40run it calibration so one of the thingsi might change hereyou can see um my peak is a little lowso one thing i might do is go in whati'm going to do is goback to components tab select the basinmodel nodeand now i'm going to globally edit myloss parametersWatch at: 30:40 / 31:00so one of the things i can do here is ican start to changemy loss rates so let's go with somethingsorry that's initial deficit[Music]Watch at: 31:00 / 31:20here we go this is what i'm looking forso i'm going to bump these to 2.75that's the wrong way 2.25okay and then i can go ahead and computefrom this menuWatch at: 31:20 / 31:40see what my graph looks like okay now mypeak's looking a little betteranother thing i might want to do isbring my base flowfocus on the recession limb and edit mybaseline parameters so i'll justglobally edit those as wellWatch at: 31:40 / 32:00i'll try groundwater 1 coefficient of 30hereand a groundwater 2 let's try 250.Watch at: 32:00 / 32:20okay let's compute thatand let's take a lookokay so now my recession ones look alittle better but i'm missing on peakflowso this is calibration we're goingthrough we're iterating we'recalibrating the model so one last thingWatch at: 32:20 / 32:39i'll doisthese are these watersheds have someurban area so i'll go ahead and add alittle bitimpervious area in them and that'lllikely bump up our peakWatch at: 32:39 / 33:00okay so we've gotten a little closer onour peak and this just demonstrates theiterative process of calibrationhow you might go about it i wouldprobably work on this some more but forthe sake of time we'll keep movingWatch at: 33:00 / 33:20so let's jump back into the slidepresentationgreat thank you tommyso hopefully you guys saw how fast youcould build a modelparameterize it input boundaryconditionssimulate and calibrate you know you canWatch at: 33:20 / 33:39always keep working on calibration untilthe end of time right but you can seehow fast you can do all that stuff withan hmshopefully you guys get a lot of benefitout of that and canget a whole lot of bang for your buckand using this uh software applicationinyour day-to-day life and your jobWatch at: 33:39 / 34:00however umnow we're going to talk about somefuture enhancements that are planned forinclusion within hms within futureversionsso first and foremost everybody'sfavorite uh we want to include the 2dflow overland flow solverspecifically from htc ras for use withinhighly detailed studies where unithydrograph transform isn't going to workfor youthat's planned for future inclusion weWatch at: 34:00 / 34:20also have simple reservoir operationsthat are planned for inclusion withinfuturereleases of hms so for instance i mightwant to simulate gated operationsat a flood control reservoir so in thissimulation herei'm showing results uh the natural flowand the dashed redlinethat is coming into the project and whati'm actually putting out of theWatch at: 34:20 / 34:40project is in blue that's realizedthrough gated operationsthrow so some logic with an hms wouldactually allow us tocompute that blue hydrograph given theinflow in redto minimize damages downstreamso another thing we're planning to workon is parameter estimation so werecently added a lot of these gisWatch at: 34:40 / 35:00features and the next stepis to do things like import a soilsdatabaseand start to analyze those databases andmake it easier for the modelers toquickly parameterize their models sothis is just an example where someone'simported a soil database we've done someprocessing on it to come up with anotion of whatWatch at: 35:00 / 35:20the constant rate parameters lossparameters might beanother area for future work is met dataprocessing so i showed youvortex and how we built a meteorologicboundary conditionbased on gridded data and we plan tocontinueworking with vortex and building outWatch at: 35:20 / 35:40that functionality like i said a lot ofit will potentially bemoved into hms as it matures anddevelops the biggest thing we need rightnowis people to test it with their datasets let us know how it goes this isactually an open sourceproject and it's available on thehydrologic engineering center's githubpage so if you want to go there you candownload it test it let me know how itWatch at: 35:40 / 36:00goesif you're inclined you can even go helpdevelop on itfinally we want to make improvements tothe existingoptimization uncertainty analysis withinhc hmsso within the current release you haveaccess to the markov chain monte carlooptimization trial and that can beWatch at: 36:00 / 36:20linkedto an uncertainty analysis toinvestigate the uncertainty as a flowor stage at a specific location solelydueto parameter uncertainty specifically wewant to makefurther reductions in runtimes to makethings faster since we have the wrongpotentially millions of simulations todo thiswe also want to add new objectivefunctions to evaluate goodness of fitWatch at: 36:20 / 36:40we also want to improve the control ofoutput variables to minimizedisk space that we need to take upso that i'll say thank you for your timefor more informationyou can visit the hydrologic engineeringcenter's web page we havethe hms page that's a good place tostart from there you can find downloadsWatch at: 36:40 / 37:00and documentationand then i want to just point at some ofthe training materials we publish soevery time we put on a class we'llpublishsome of the training materials and youmight find that helpful to get startedon your projectsso thank you everybody for spending thetime again truly truly appreciate theopportunity to presenthsc hms to you all thanks tom thanksmichael it'sWatch at: 37:00 / 37:20it's been a total feast um simulatinghydrologic processesit's been fantastic we can see theenormous amount of experience behind itthank you everyone for your questionsand you'rei i haven't broken into the the the umthepresentation simply because it was justso gripping and engagingand it was good to get the whole pictureuh craig can i hand back to you andWatch at: 37:20 / 37:40sorry but that went it went deadwhat chances are there that that are anoutage should happen right in the twominutes when you'redoing your presence i know exactlyexactly can you hear me okay now thoughvery clearvery okay good um it's it's back on yeahsomebody at the street cut the powerright in the middleof my talk so that that's fine um i sawyou comingright with something yes yes so umWatch at: 37:40 / 38:00hopefully uhum uh joel will work his magic ontrimming up the video for uh for youtubeviewers but yeah that i i did get to getto see most of that um and veryinteresting i've been trying to answeras many questions as we canthere are a few actually still on therethat um uhsteve and i are still getting to so tomWatch at: 38:00 / 38:20and michael feel free to uh pop in thereandhit the q a as well while we're in thesession live you could also have a lookat some of the text ones we'll we'll tryto address a few of thoseand um and i think a lot of thequestions that were coming inum as the questions came in you guyscovered that um as far as the the futureworkWatch at: 38:20 / 38:40so so all right we can probably diveright into those let'ssee i'll the way we'll do this i'll readthem out because the recording doesn'tactually see the questions ifwe pick the questions that have beenupvoted i'll read them out and we canget stuck into thembefore i start um you can see how manypeople are on this webinaracross the world it's going to be afairly full class there's only 30so i'm not doing a big sell here but iWatch at: 38:40 / 39:00probably am but umbut if you want to get in on this courseand i can only say start registeringas soon as the webinar finishesgentlemen do it i think uh joel yeahthank you for thatuh joel's put a registration up on thechat line there you'll see everybodyuh steve steven craig will be runningthat course butwith tom and michael's backing has beenfantastic let's get goingWatch at: 39:00 / 39:20uh first question from professorfrancisco in mexicois there any perspective to the heck hmsto be an open source in futureand is there any risk that it won't beavailable for free in the futuregood question okay sothis is uh i guess i can speak to thatWatch at: 39:20 / 39:40i don't see hms being open sourcein the new near future i do seeing itbeing free uh forfor the foreseeable future so one of themost difficult things working here isyeah we're tied in with dodthere's some restrictions there butthere's alsojust our capacity at hcc a lot of peopleWatch at: 39:40 / 40:00think we're this big softwaredevelopment firm of300 or 1000 but it's really the hmsteam is five to six people i would sayand so open source is great and andlittle projects like vortex are opensource so weWatch at: 40:00 / 40:20we're open sourcing bits and pieces butthe ability to manage a fullopen source project we would truly needmore bodiesto appropriately manage that but yeahfor the foreseeable futurehms is free it's been kind of one of thefounding principles is to distributethe free software so yep yep that'sWatch at: 40:20 / 40:40right you crack a lot of uhyou you you hit well above your weightwithout only half a dozen peopleeverybody everybody's very keen about ituh right should we go on the nextquestionuh tahira is asked what's the differencedifference between heck what heckhms those two and applications of heckwhat versushms who wants to take that one on soWatch at: 40:40 / 41:00hcc watt thank you this is an awesomequestion and this isvery very cool my buddy will willdefinitely love this who's the team leadof hcc watthtc watt stands for the watershedanalysis tool so this is actually one ofour integration pieces of softwareso you saw earlier a slide where iincluded ascreen capture from our core watermanagement system swimsWatch at: 41:00 / 41:20this is very similar to that but for aplanning mode so htc wattallows for the integration of hcc hmshsc resim agency raz and hccfia to dothe fullanalysis suite of hydrologic simulationreservoir analysis simulationhydraulic simulation and consequencesWatch at: 41:20 / 41:40estimation as well all in one fell swoopso htc watt allows you to donot just deterministic computes whereyou know exactly what everything is butalso stochastic computesso we can sample from the uncertaintyspace for parameter inputsfor boundary conditions and also fullyanalyze uncertaintyuh all the way through to consequencesWatch at: 41:40 / 42:00so that is a very very cool questionhsc hms is a component model of hcc whator can be a component model of hc1awesome questionuh thanks a lot and thanks for thatquestion here uh where do we go nextcrowd with these questions what do youthink uhjoel please question yeah well one onequestion that seems to have come up in acouple of different forms and uhWatch at: 42:00 / 42:20this this is one that i think you guyswill have a standard answer for when isuh 4.4 coming out uh and and i know uhyou know on on the hec razz side garyalways tells us not tonot to even ask and even if he answersdon't quote him on it soum yeah when might 4.4 be expectedyeah so if you want to get your hands onWatch at: 42:20 / 42:4044 today andhelp us test it we have a beta testingprogram so the way to do that will be toemailthe hms inbox you can find it on the hmswebpagei would try to say it but i wouldn'tquote me on it i believe it's hmsusac dot army dot mil but except theatc.hmsWatch at: 42:40 / 43:00anyhow find that emailsend an email to that requesting 44 betauh that'll eventually get to me one wayor anotherand then i will send you anon-disclosure agreement which justkind of tells you what's appropriate fortesting the data software like don't doany official studies with this becauseit's not officially releasedWatch at: 43:00 / 43:20things like that and if you agree to thenon-disclosure agreement we send you thetestsoftware and you can start to test withit uh formal release is hard to say likecraigmentioned gary's gary's very dodgy aboutputting down firm dates and we'velearned that in the software developmentworldyou can never be too sure so i iWatch at: 43:20 / 43:40yeah i would anticipate uh how can i seethis generally soon as we canas soon as we can yeah let's go withthat yeahgreat thank you excellent okay welljalva's question is can thehydrologically corrected data beimported from arcmap who wants to take that one onyes so we do have plans for that pleaseWatch at: 43:40 / 44:00view these gis tools asalpha versions right so this is ourfirst rollout ofintegrating gis functionality in aholistic sense within hms we are goingto make things betterin the future we're going to continue tomake improvements soif you have suggestions like this orthings that you wish it would dochances are we probably are going afterWatch at: 44:00 / 44:20that in the future but we would love tohear from youso please send that to the hms inbox aswellyeah and keep that in mind going forwardi'll add to what mike says with theso uh on the meteorologic side all thosedss records we have the ability there'savortex utility to export those metWatch at: 44:20 / 44:40records to geotefso you can export to geotef and view inany gisthe gis features in hms those live in asqlite databaseinside of the project folder so if youpoke around in there you can find thesqlite databaseyou probably have trouble viewing thoseyou have to do some special importingWatch at: 44:40 / 45:00steps to view those in esri productsif you're using qgis you can drag anddrop and they open right therewell should we keep moving yeah i meanthere's there's quite a fewthat we're not going to be able to getthrough all of these so um mike and uhand tommy if you want to again just ifyou see any of these on there that youWatch at: 45:00 / 45:20could uhanswer quickly or wanted to highlightjust have a have a brief look at itbecause we need to kind of pick andchoose which ones toum to handle and uh yeah i mean a lot ofpeople have asked about the integrationwith hecrazwhich i think you've already covered umand uh when that might be happeninguh one one that i thought um might beworth uh covering now theWatch at: 45:20 / 45:40the an api for uh web applications isthereuh is that something that's on thehorizonyeah i'll talk about api a little bitthat is on the horizoni guess i should clarify too all of thehms development you know a lot of peoplethink oh we're us government we get abig pot of money and wego to town with it which isn't actuallythe case we work closely withWatch at: 45:40 / 46:00development partners who have specificgoals in mind and sowe usually partner with them throughthings like memorandums ofmousse memorandum of understanding sothat's how we've partnered withother countries and things like that sothe apiwe actually have a few developmentpartners that are keen on thistechnologyWatch at: 46:00 / 46:20and this this capability we've had akind of a course unpublished api in thepastso that functionality is there there'ssome basic documentation in the user'smanualbut one of the things that we'll beworking on in this next week actuallytoday is thethe beginning of our fiscal year so wejust kind of resetWatch at: 46:20 / 46:40the books and we're working on newfeatures and so one of the things wehave on tap for this next fiscal year isthat apiso if it's not and it likely won't be infour or fourbut uh in subsequent releases you'llprobably see some basicapi functionality and with that mostimportantly is the documentation on howWatch at: 46:40 / 47:00tointeract with the apiexcellent um so yeah as i guessi'm trevor i'll just uh keep having alook at some of these ones here umand and just kind of group them into uhinto groups and again feel free um steveoruh mike or tommy to answer these butagain uhthose who are not on this session liveWatch at: 47:00 / 47:20and we'll be watching on on youtubeum won't see these questions so we canuhwe can kind of handle uh handle this oneumlive so that they can see the hear theanswers right now umdoes hcc have a tool for rainfalldisaggregation umthat's i think i think you may havecovered part of that but um want toaddress thatWatch at: 47:20 / 47:40uh yeah so uh you can do that with anhmsso you can input a daily time series ofprecipitation and runat a one minute time step if you wouldlike is the smallest time step that youcan usehowever um that's still not going toapply a specific patternso it's just going to interpolate fromthere so if you're specifically askingWatch at: 47:40 / 48:00if you can apply a pattern to thatum that is something that we haven'tquite done yet though that is a greatquestion and a great thing to tackle inthe futureexcellent yeah i mean it sounds likethere's a lot of things coming up uh inthe in the future that umuh that we could look forward to um wemay be lookingwell into the future for some of thesequestions that have come up about theWatch at: 48:00 / 48:20integration some people have asked umare we going to be able to get this torun together with groundwaterand with sediment transport and you knowsediment yield and i would love to havea grand unified model where i can justopen it up and in one fell swoop uh geteverything modeled all at onceum for now i think we're still going tobe stuck with different models and usingboundary conditions from oneWatch at: 48:20 / 48:40to feed into the other and quite amanual processand you know in the end we are stuckwith models and all these models arewrongmaybe i'll cover a quick one minute ofwhat was going to go into myintro really quickly because i i put ibrought some toys along here as wellto show uh maybe i'll i'll share myWatch at: 48:40 / 49:00screen real quick just to show you whatthese uh some of theseother models might be hopefullyi hate to do this during the q a partbut the question has come up a number oftimes are you able to see my screenthereabsolutely this is where we left offwith the prequel and i will just uhgo heading back in time and covering theWatch at: 49:00 / 49:20integration between heck hms and hec-rasthat question came up a few timesand just realizing that hydrology whenwe're doing uh you know thequantification of the flow and gettingthese hydrographsand the characterization of that flow ina hydraulic modelthat there is a bunch of overlap whatcomes into the hydrological modelshas to be routed using open channel flowWatch at: 49:20 / 49:40methods and nowwith hecarius and other software you canuse grain on gridand use that as a hydrological tool butthey ought to becalibrated against each other just sothat you get reasonable results umeven when you're using these rain ongrid results now a model itself um thisguy isnot the model he's got a model here thatWatch at: 49:40 / 50:00you can seei've got a few of these as well in frontof me that i use as classroom toolsand um in in some of these ones um iguess i just wanted to show you a coupleof theseum as as as kind of video displays andi'll put some of these upand and you can you can actually haveyour fifth grade projectWatch at: 50:00 / 50:20um if you want looking for a scienceproject uhthis is actually a very easy one youdrop the rain inand uh you have a quantity of rainfallgoing in and you see what the runoff iscoming outand a lot of these models are used as uhexamples of how much sediment comes outand i wish we could do that all at onceanddetermine all the sediment yield and theWatch at: 50:20 / 50:40sediment transport and the scourum but for now it's it's kind ofdifferent different modelsum but what i wanted to do here is justagain show you a couple ofthe ones that um that i've got here i'lljust uhstop sharing my screen here umjust losing your face a bit there ohWatch at: 50:40 / 51:00yeah i'm i'm gonna actually point thisdownnow and show you the ones that i've gothere um just so that we can get thedifference betweenuh heck as a rainfall uh a rain on gridmodeland um and uh the the heck hmsand so i don't know if you can see thisthing right here i always use these toysum i'm gonna have this digital elevationWatch at: 51:00 / 51:20modelthat i'll build a little river andwatershed into that's what you've got ina rain on grid modelwhat i do in these ones then is put thatinto my basinright here and i overlay it with acomputational gridwhich i'm going to put right over thetop of it and that computational gridthenwill get rain on every single one ofWatch at: 51:20 / 51:40these grids so i've got this little toythat i use here which gets thesesyringesand as i push this thing down right nowin heck razzyou can change the tempo of this i canpush down fast and push down slow andsquirt water into each one of thesegridsi can't as yet change the temporalpattern and push one to get a stormWatch at: 51:40 / 52:00moving one direction or another that'scoming in the next versionbut what we've got then with rain goingon every single one of these grids inhecara'scompare that to hms which is like thosebottle models that i showed you and i'llput those up hereas well when you drop water into this uhdirt basically have dirt and watercoming together and you measure how muchWatch at: 52:00 / 52:20is coming out we're quantifying the flowgoing in and we're quantifying the flowgoing outbased on these parameters soil lossescurve numbers um infiltration rates andrunoff coefficients this this ismathematically basedversus a rain on grid model which manyof you have attended thoserain on grid webinars in the past umWatch at: 52:20 / 52:40that is just looking at the topographyand we try to slow the uh slow the waterdown with roughness and we try to do allthese things that don't quite simulatethe same thingyou can get the same results out you cancalibrate them to be giving you the sameresultsbut they're two completely differentphysical mechanisms and having thesemodels talk to each otherWatch at: 52:40 / 53:00is a very challenging process but youcan do it either way you can do it asrain on grid or you can do it as arainfallrunoff routing model and you can getresults outbut what we need to realize is we don'tactually know the resultsthe those results are are theuncertainties that go into the rainfalland the uncertainties that come out ofwhat's going on in the soilWatch at: 53:00 / 53:20are still very very great and so we'regetting better and better at modeling itand getting these temporal patterns andspatialspatially variable patterns coming infrom radar data and everything else butthere still is a lot of uncertainty inthese models and calibration data iskey absolutely key when you work in anarea that has no calibration data justWatch at: 53:20 / 53:40remember your model is probably wrong itactually is wrongand uh you know we we do our best to getit as right as we canbut it's never going to be absolute soanyway that was my lengthy littleexplanation on a few of these questionsthat i've already already answered thatwere betweenuh comparing rain on grid in hecara'sversus rainfall runoff uh modelingWatch at: 53:40 / 54:00in hms and uh obviously marmike mike and tommy have done a muchbetter job of explaining the hms pieceof thatbut for our hec-ras fans who have tunedin on this one i just wanted to makesure we highlighted those differencesthank youquestions then i'll go ahead and putsome comments here just let me make somecommentsWatch at: 54:00 / 54:20stephen hi umbeen a great explanation from michaeland tomand really top-end a lot of the work idois a lot simpler than that a lot ofsmall projects a lot of flood studiesand when i do my course live coursesas i will do in the webinar in the inWatch at: 54:20 / 54:40the future coursewe'll really get down to basics and keepthings simpleso you can build up from you know twoplus one plus one equals twoand two plus two equals four and andwhat we've seen this this um thismorning or this evening depending whereyou are in the worldwe can you can see the potential of whatWatch at: 54:40 / 55:00it can dowhen you've got the technology and databut at the end of the day if you don'thave the data it's going to be very hardto build the models buti can build a lot of models and getquite goodflows to put in for example a necrozmodelthat will enable you to do veryeasy jobs very quickly andWatch at: 55:00 / 55:20i know that michael and tom like to dothe big projectsum massive square kilometers where for alot of engineersand i could be wrong that could be it inthe wrong place hereare dealing with very small catchmentstrying to designpipe systems trying to understandflooding of streams where they'reWatch at: 55:20 / 55:40buildingum commercial properties for example somy comment is i'll pare it down to be alot easier for people to understand whenit comes to the coursenow that sounds great and look um we'regonna have to wrap this up here we'vejust hit the hour mark and it's beenwonderful having your time gentlemen andeveryone on boarduh with all your questions we haven't uhWatch at: 55:40 / 56:00hit all of them i understand that umbut i do appreciate that we've got thehigh sciencewe've got the high modeling and we'vegot the high practice ohand the stuff you were showing just amoment ago craig just makes it soso uh obvious to everyone and and likeyou say steveum it's the it's the large uh broad areaWatch at: 56:00 / 56:20uh right through to the smaller umacreagesyou know that that need to be managedherelook why don't we have a a a quick goaround of a last word from each oneuh on this topic then we'll close downthe webinarand i'll invite um each one of you nowjust to have a bit of aWatch at: 56:20 / 56:40one-tenth or two-sentence um wrap-up umwho wants to kick offor well maybe we'll kick off with ourguests michael and tomokay yeah i'd just like to say thanksfor this opportunityand i'm enthused whenmodelers and users do really cool stuffso if you do cool stuffuh let us know about it send it to usWatch at: 56:40 / 57:00and feel free to reach out and contactus if you have any questions with thesoftwareanything like that thank you so much forthe opportunityto discuss today i'm super super happyto seeeverybody from around the world it'salways so flattering andawesome to see people using uh oursoftware all over the place and all overWatch at: 57:00 / 57:20the world please send us some use caseslet us know how you're using ityep and yourself steveme next yep okay and now it's beenreally goodyou've got to remember as a modeler thatum hms like any tool is only atall and you've always got to have yourbackup we're engineers or scientistsWatch at: 57:20 / 57:40we've always got to know and understandwhat we're doing we don't throw numbersinand it's a great way of doing like i sayi use it every day practicallyand it's a great way of producing somehydrographs and volumes of dischargeinto basins etcand but you've always got to have abackup you've always got to understandWatch at: 57:40 / 58:00what you're doing i like that it takesthe dark art out of it it's not a blackbox it's actually got real data andit's got it you've got to have realground ground truthing uh and yourselfcrayoh yeah um i don't want to take too muchmore time just uh but i don't want toleave anybody hanging everybody who'sgot these questionsand uh asked here um even those who thathave been answered anduh dismissed um if we can just collectWatch at: 58:00 / 58:20thoseand then give tom and michael a chanceto weigh in on thoseand then we can respond individually sowe don't want to leave anybody hangingif you asked a questionuh we do have it recorded we um on ourends and so we'll make sure we get youtaken care ofi'll show you yeah if you're able to dothat that'd begreat absolutely yep no thank you forthat that's that's been really goodWatch at: 58:20 / 58:40gentlemenum and as craig said i'll i'llcollate those questions and shoot themout to every one of the presenters uh soyou will hear from usuh well thanks for participating thisthis feedback going to appearin a moment uh on your screen uh therecording will email to you uh there'sthe threethe webinars coming up in the next fourweeks uh and online courses pleaseWatch at: 58:40 / 59:00please do join us for the umuh the hms course coming up in novemberthere's also one as you can see there inoctober on hec rasand others there i won't go through eachone go to our website golook on the youtube and you'll have allthe details once againtom michael steve craig it's been afeast it's been a fantastic hour weWatch at: 59:00 / 59:20should go on for another two or threeit's absolutely brilliant appreciate allyour efforts and timeso for now it'll be a goodbye toyourselves and to everybody watchinghere today thanks very muchbye for nowsubscribe to this youtube channel byclicking the subscription buttonWatch at: 59:20 / 59:40for future webinars and online shortcourses please visit our websiteat australianwaterschool.com

About The Author

You Might Be Interested In

Comment (0)

  1. area to define streams depends on what? on what basis did u enter 70 sqkm for area to define streams? Is it the basin area? If at all we are taking it as a basin area we are facing this error WARNING 46501: Unprojected coordinate systems are not recommended for the basin model. as river basin huge its area is 2,58,948 sqkm

  2. Very helpful video, but when I run simulation for my project I get an error message (21401) saying there are no elevation bands set for subbasin. Does anyone know why this is happening, or more importantly how I can fix it? There doesn't seem to be anywhere in HMS where I can set elevation bands, as far as I know. My basin model is of course linked to a DEM, shouldn't that be enough?


Your email address will not be published. Required fields are marked *