Hi, I'm an error. x

Data For Terrain Analysis Enhancements to the Height Above Nearest Drainage Flood Inundation Mapping Method


Authors: Irene Garousi-Nejad David Tarboton Mahyar Aboutalebi Alfonso Faustino Torres-Rua
Owners:
Resource type: Composite Resource
Storage: The size of this resource is 18.0 GB
Created: Dec 31, 2018 at 5:38 p.m.
Last updated: Jun 28, 2019 at 5:54 p.m.
Citation: See how to cite this resource
Content types: File Set Content  Geographic Feature Content  Geographic Raster Content 
Sharing Status: Public
Views: 242
Downloads: 3
+1 Votes: Be the first one to  +1 this.  (You need to be logged in to rate this.)
Comments: No comments (yet)

Abstract

This resource contains the input/output and scripts used for Terrain Analysis Enhancements to the Height Above Nearest Drainage Flood Inundation Mapping Method research which is submitted to the peer-reviewed journal of Water Resources Research and is not formally published on HydroShare in case any change are needed during the review process. Once the paper is accepted, a DOI will be used to cite this resource in the paper.

The abstract from the paper follows:
Flood inundation remains challenging to map, model, and forecast because it requires detailed representations of hydrologic and hydraulic processes. Recently, Continental-Scale Flood Inundation Mapping (CFIM), an empirical approach with fewer data demands, has been suggested. This approach uses National Water Model forecast discharge with Height Above Nearest Drainage (HAND) calculated from a digital elevation model to approximate reach-averaged hydraulic properties, estimate a synthetic rating curve, and map near real-time flood inundation from stage. In 2017, rapid snowmelt resulted in a record flood on the Bear River in Utah, USA. In this study, we evaluated the CFIM method over the river section where this flooding occurred. We compared modeled flood inundation with the flood inundation observed in high-resolution Planet RapidEye satellite imagery. Differences were attributed to discrepancies between observed and forecast discharges but also notably due to shortcomings in the derivation of HAND from National Elevation Dataset as implemented in CFIM, and possibly due to sub optimal hydraulic roughness parameter. Examining these differences highlights limitations in the HAND terrain analysis methodology. We present a set of improvements developed to overcome some limitations and advance CFIM outcomes. These include conditioning the topography using high-resolution hydrography, dispersing nodes used to subdivide the river into reaches and catchments, and using a high-resolution digital elevation model. We also suggest an approach to obtain a reach specific Manning’s n from observed inundation and validated improvements for the flood of March 2019 in the Ocheyedan River, Iowa. The methods developed have the potential to improve CFIM.

There are four main directories in this resource, each of which is described in details in Readme.md, which we recommend reading the Readme.md prior to running scripts to reproduce results.
To reproduce results:
1- Select which scenario you want to run (more details on scenarios can be found in the paper)
2- Download the directory of the selected scenario from this HydroShare resource (for example ./Processed_Data/Scenario6). This contains required inputs (including terrain analysis and HAND rasters) to run python scripts. If one also wants to reproduce including terrain analysis and HAND rasters, we recommend obtaining them as described in the methodology section of the paper.
3- Download the directory including scripts (./Scripts). Note that some functions used in these scripts require the arcpy library. In addition, note that addresses and inputs defined in scripts are based on Scenario 6. Scenario 6 is the scenario where we applied our developed etching approach to the 3(m) DEM using the high-resolution hydrography dataset along with all other improvements evaluated in scenarios 2-5. If using these scripts with another scenario, one needs to change addresses and name of inputs.
4- The easiest way to start working with these scenarios and scripts, is to do step 2 and 3 above and put them in the following address on your local machine: C:\Scenario6\
5- You can evaluate reproduced results with those we put into ./Output_Data directory of this HydroShare resource.

Subject Keywords

  • No subject keywords have been added.
  • ${ k }

Duplicate. Keyword not added.

Error: ${ error }
Deleting all keywords will set the resource sharing status to private.

Resource Level Coverage

Spatial

Coordinate System/Geographic Projection:
WGS 84 EPSG:4326
Coordinate Units:
Decimal degrees
North Latitude
41.8419°
East Longitude
-112.0371°
South Latitude
41.5583°
West Longitude
-112.1418°

Content

Readme.md

This resource contains the raw data, processed data, scripts, and the outputs of the: Terrain Analysis Enhancements to the Height Above Nearest Drainage Flood Inundation Mapping Method.

Base_Data:

This directory contains the raw data used for this research.

  • DigitalElevationModels:
    Folder holding 10 and 3 meter (m) National Elevation Dataset Digital Elevation Models (DEMs) for the study area obtained from the national map viewer.

  • ModeledStreamflow:
    Folder containing the National Water Model analysis and assimilation hourly (as used in scenario 1 in order to be the same as CFIM method) and average daily (as we used to create Figure 4 of the paper) streamflow for National Hydrography Dataset (NHDPlus) reaches within the study domain (i.e., the same as the domain for which the 3 (m) DEM is availabe) obtained from National Oceanic and Atmospheric Administration (NOAA). The script that we used to download these data is provided under the Scripts/NationalWaterModel directory of the current HydroShare resource.

  • ObservedStreamflow:
    Folder holding the historical and observed 15-minutes, daily average, and annual peak streamflow data at two gages (i.e., PacifiCorp at Collinston and United States Geological Survey (USGS) gage 10126000 at Corrine) on the flood date obtained from Bear River Commission and the National Water Information System of the USGS, respectively.

  • MediumResolutionNHDPlus:
    Subset of the medium resolution NHDPlus reaches for the Hydrologic Unit Code (HUC2=16) that includes this study area obtained from this HydroShare resource. This is a subset prepared for the National Flood Interoperability Experiment (NFIE) and has reaches for which the National Water model forecasts are not produced, excluded [Predict no NWM results for missing flowlines].

  • MediumResolutionNHDPlusCatch:
    Subset of the medium resolution NHDPlus catchments the HUC2=16 that includes this study area obtained from this HydroShare resource.

    • HighResolutionNHDPlus:
      The high resolution NHDPlus reaches for the HUC4=1601 that includes this study area obtained from the national map.

    • shapefile_3mdemdomain: The shapefile of the study domain derived from the 3 (m) DEM extent.

    • shapefile_gages: The location of two streamflow gages (PacifiCorp at Collinstone and USGS gage 10126000 at Corrine).

Processed_Data:

There are six directories related to each scenario. In each scenario, there are four directories including the processed data that we made ready so that to be used later in scripts, which we developed to generate results for this study.

  • Basic:
    This directory contains the shapefile of each catchment (NHDPlus-based or DEM-derived depending on the scenario) within the study domain draining into the reach of Bear River between two steamflow gages (PacifiCorp at Collinstone and USGS gage 10126000 at Corrine). It also has the shapefile of the study domain (called 3mdemdomain.shp) and the shapefile of the domain we used to compute the evaluation metrics (called domain_Metric.shp). There is also another shapefile which is the merged shapefile of the catchmetns (catch_mrg.shp) and its corresponding raster file (catch_mrg_rs). Moreover, the folder contains some text and comma-seperated values (csv) files, which are used later in the developed scripts. Here we introduce each one of these text/csv files:

    • Adjusted_Q.txt: This is the daily average observed streamflow values for each reach (NHDPlus-based or DEM-derived depending on the scenario) along the Bear River between two gages (PacifiCorp at Collinstone and USGS gage 10126000 at Corrine). This is used when we want to create the flood inudnation map based on the observed flood discharge. Note that this text file is only availabel for scenarios 2-6. Scenario 1 does not have this text file because this is the base scenario and everything is kept the same as being defined in the current CFIM method. In scenario 1, the hourly streamflow values are the National Water Model analysis and assimilation product. Please refer to the paper for details. IN scenario 1, the text file including the National Water Model discharge is called NWM_Q.txt.

    • INDEX.txt: This text file contains a column of integer values (we call index values) which are used within the developed script, called 03_script_03_Hydroprop.py, when generating proper names for files. Note that the values in this text file do not have any physical meaning. This file is not availabe for scenario 1, because preliminary results of scenario 1 were obtained while the author was attending the University Consortium for Geographical Information Science (USGIS) summer school in 2017, supported by the National Science Foundation (NSF).

    • linkid.txt: This text file includes the information about the reach/catchment id, the reach slope (adjusted slope from our slope adjustment approach), the reach length, and the average latitude of the catchment. These information are required when creating the hydraulic properties table.

    • STAGES.txt: A text file containing stage values (ranging from 0.1-27 meter) that are used to create flood inundation maps using HAND method based on which the hydraulic properties and synthetic rating curves are generated. This is not available for scenario 1, becasue this is the base scenario and everything is the same as CFIM and we generated results (including hydraulic properties and rating cureves) while the author was attending the University Consortium for Geographical Information Science (USGIS) summer school in 2017, supported by the National Science Foundation (NSF).

  • Input_Raster: This directory contains two raster files [(Height Above Nearest Drainage (HAND and D-infinity slope (slp)] required to compute the hydraulic properties and flood inudnation map. Scenario 1, contains two more raster files (dem.tif which is the DEM raster and inlets.tif which is the raster of the NHDPlus medium resolution starting points). We just put these in here in case one wants to see which inputs are used in CFIM to generate HAND.tif and slp.tif files.

  • Planet_Observed: This is the classified observed flood inudnation extent. We used the supervised classification method to create this map from Planet RapidEye satellite image on the flood date. In each scenario, te spatial resoltuion should be the same as the DEM and HAND rasters of that specific scenario.

  • Slope_adjustment: Results of using the slope adjustment approach. Slope adjustment approach is developed because in some cases we got slope of zero after terrain analysis processes for reaches that were across flat areas, and these became artifacts in the DEM. In these cases, we adjusted the local elevations of the junctions in the digital representation of the stream network to shift elevation changes between stream reaches and impose a non-zero slope on each stream reach. For example, when the reach downstream of a junction has a positive slope, but the reach upstream is flat, the elevation of the junction can be lowered to make these reaches have an equal but smaller slope than the downstream reach did initially. We extended this idea both upstream and downstream to accounting for the occurrence of adjacent reaches with zero slope. The result was a set of reach slopes that are all positive but do not alter the, to equalize upstream and downstream slopes without altering overall elevation differences and hence slopes in the stream network. For more detailed information visit this. Note that Scenario 1 and 2 does not have this folder, because NHDPlus slope values are used in these scenarios, not the slopes computed from terrain analysis processes.

Scripts:

This directory contains 13 python scripts that can be used to reproduce results of this study. Note that these scripts should be run in the order as that they are numbered. Prior to running these scripts, obtain the terrain analysis (TauDEM functions) and Height Above Nearest Drainage (HAND) rasters as described in the methodology section of the paper. In addition, note that some functions used in these scripts require the arcpy library. The addresses and the name of inputs are based on those used in Scenario 6. Scenario 6 is the scenario where we applied our developed etching approach to the 3(m) DEM using the high resolution hydrography dataset along with all other improvements evaluated in scenarios 2-5. If using these scripts, one needs to change addresses and name of inputs.

  • 01_script_01_Hydroprop.py:
    This script takes the HAND raster and STAGE.txt (a text file of different stage values) as inputs and generates a raster with all grid cells less than a specific stage value for each stage value in STAGE.txt. It also creates a raster indicating flood inundation depth at each grid cell using the HAND value (from the HAND raster) and a specific stage value for each stage value in STAGE.txt.

  • 02_script_02_Hydroprop.py:
    This script takes: (1) D-infinity slope raster, (2) the "Less Than" and "Depth" rasters output from the "01_script_01_hydroprop.py", and (3) the "linkid.txt" (a text file including the id of each reach or catchment). Results of this code are three csv files for each catchment. These csv files are: (1) the average inundation depth (called depths.csv), (2) the number of flooding cells (called cells.csv), and (3) the square root of 1+(D-infinity slope)^2 (called srp.csv). These csv files are then input to "03_script_03_hydroprop.py" to create the base hydraulic properties table (i.e., river geometry).

  • 03_script_03_Hydroprop.py:
    This script calculates the river geometry using all cell.csv, depth.csv, and srp.csv files (output from "02_script_02_Hydroprop.py").

  • 04_script_04_Hydroprop.py This script uses the river geometry csv file (the result of the "03_script_03_Hydroprop.py") and computes reach-averaged hydraulic properties (called full hydraulic table) and synthetic rating curve for each catchment.

  • 05_script_EstimateFloodStage.py:
    This script uses the observed discharge values and the synthetic rating curve (created from the script above) to estimate the flood stage for each reach/catchment.

  • 06_script_Inundationmap.py:
    This script uses the estimated flood stage and the HAND raster to map inundation. Results are inundation maps for catchment. These are then merged in "07_script_Confusionmatrix.py" as one complete map that shows the HAND based flood inundation extent.

  • 07_script_Confusionmatrix.py:
    This script uses the classified observed inundation from satellite and the HAND based flood inundation map to compare modeled and observed inundation by creating the map of the confusion matrix. This map is later used to compute evaluation metrics Fit(F) and Corectness (C). These metrics indicate the degree-of-overlap between model and observed flood inundation maps. Both metrics should ideally be 1 (100%). C is an overall area metric and F is a location- specific metric

  • 08_script_Evaluationmetrics.py: This script computes evaluation metrics (i.e,. Fit, F and Correctness, C) for the entire domain.

  • 09_scriptEvaluationmetrics_over_specificCATCH.py:
    This script computes evaluation metrics (i.e,. Fit, F and Correctness, C) for each catchment.

  • 10_script_FandCforStages.py: This script is the first step of the optimization procedure (scripts 10-13) whereby the optimal stage value is defined. This script computes hydraulic properties, synthetic rating curves, flood inundation stages, flood inundation maps, and evaluation metrics for a range of stage values for a specific catchment. Remember to define the catchment in the "Define the following parameters" section of the script.

  • 11_script_BestStageBasedonBestF.py: This script finds the optimal stage value based on the maximum value of F (computed from results of the script above) for each catchment.

  • 12_script_BestInundationdepthBasedonBestF.py: This code finds the optimal average flood inundation depth value based on the optimal stage value based on linear interpolation of the hydraulic table. This value is needed to back calcualte an optimal Manning's n.

  • 13_script_BestManningBasedonBestF.py: This script identifies the optimal Manning's n based on the optimal average flood inundation depth.

Output_Data:

This directory holds results of this study. Under this directory, there are six directories that are related to each scenario. For Scenarios 1-5, there are four directories including outputs of the terrain analysis using TauDEM to create HAND map and results of the above scripts. For Scenario 6, in addition to four directories such as other scenarios, there is one more directory that includes results of the optimization approach used to estimate local optimized values for Manning's n. Note that we do not provide the temporary results (created during the run-time processes) on HydroShare because these temp files are not needed for final analyses.

  • Output_TauDEM_HAND:
    This directory contains resutls (feature and raster files) of the TauDEM and GIS functions.

    • Feature: This includes several shapefiles:

      • Catchpoly: Dem-derived catchments

      • drainageline: Dem-derived streams

      • Evenly_dispersed_nodes: Nodes that are used to make evenly uniform reaches and catchments. Note that is is only available for scenarios 3-6. Because this is one of the developed imrpvements to CFIM and our improvements are used and evaluated through scenarios 3-6.

      • NHDFlowline_HR_ArtificialPath: The artifical path from high resolution NHDPlus flowlines. This is used as the high resolution hydrography dataset in the etching process. This is available in scenario 4 and 6 where we applied etching method.

      • NHDFlowline_HR_PipelineConnector: The pipeline and connector segments from high resolution NHDPlus flowlines. This is used as the high resolution hydrography dataset in the etching process. This is available in scenario 4 and 6 where we applied etching method.

      • NHDFlowline_HR_StreamRiver: The stream river segments from high resolution NHDPlus flowlines. This is used as the high resolution hydrography dataset in the etching process. This is available in scenario 4 and 6 where we applied etching method.

    • Raster: This includes several tif files:

      • dem: The oroginal DEM used in each scenario.

      • etched_dem: The etched DEM after applying the etching (flow direction conditioning) approach. Only available in scenario 4 and 6.

      • hand: The Height Above Nearest Drainage raster file.

      • inlets_on_mainstem: The raster of head (start) points of the medium resolution NHDPlus flowlines that is used as starting points when delineating the stream network from DEM using D8 method.

      • slp: The D-infinity slope.

      • srfv_artificialpath: The raster of artifical paths used in the burning process. Only available in scenario 4 and 6.

      • srfv_pipelineconnector: The raster of pipe lines and connectors used in the burning process. Only available in scenario 4 and 6.

      • srfv_streamriver: The raster of stream rivers used in the burning process. Only available in scenario 4 and 6.

      • zb: The burned dem. Only available in scenario 4 and 6.

      • zbfel: The pit removed burned dem. Only available in scenario 4 and 6.

      • zbp: The D-eight flow direction based on the burned dem. Only available in scenario 4 and 6.

      • zbpm: The masked flow direction based on the burned dem. Only available in scenario 4 and 6.

      • zbsd8: The D-eight slope of the burned dem. Only available in scenario 4 and 6.

  • Output_Hydroprop:
    There are four csv files within this directory:

    • hydroprop_base.csv: The base hydraulic properties (i.e., river geometry) information for each catchment based on several stage values.

    • hydroprop_full.csv: The compelete hydraulic properties table using the defaul Manning's n value as used in CFIM approach (n = 0.05).

    • hydroprop_full_nOptimal: The compelete hydraulic properties table using the optimal local Manning's n values which are computed using script number 12. Note that this is only available for Scenario 6.

    • hydroprop_full_n0018299: The compelete hydraulic properties table using the average of the optimal local Manning's n (n=0.02).

  • Output_Inundation:
    Resuts of the inundaion map. There are three sub directories each of which corresponds to a different Manning's value (i.e., default, optimal local, and the average of optimal local). For scenarios 1-5, there is only one directory which refers to the defaul Manning's n value (n=0.05) being used in that scenario.

    • n005: When Manning's n is 0.05

      • confusion_metric_raster: The raster that is the result of observed flood extent minus modeled flood extents. This raster is then used to compute evaluation metrics (C, the correctness, and F, the fit).

      • hand_inundation_extent: The flood inundation extent based on HAND method.

      • observed_inundation_extent: The flood inudnation extent from the classified observed flood from Planet CubeSat satellite.

      • InterpolatedStage.csv: The estimated flood stages from the synthetic rating curves based on the flood discharge (observed or modeled). This is the result of script number 5.

    • nOptimal: When the optimal local Manning's n values are used.

      • confusion_metric_raster: The raster that is the result of observed flood extent minus modeled flood extents. This raster is then used to compute evaluation metrics (C, the correctness, and F, the fit).

      • hand_inundation_extent: The flood inundation extent based on HAND method.

      • observed_inundation_extent: The flood inudnation extent from the classified observed flood from Planet CubeSat satellite.

      • InterpolatedStage.csv: The estimated flood stages from the synthetic rating curves based on the flood discharge (observed or modeled). This is the result of script number 5.

    • n0018299: When Manning's n is 0.02, which is the average optimal local Manning's n values.

      • confusion_metric_raster: The raster that is the result of observed flood extent minus modeled flood extents. This raster is then used to compute evaluation metrics (C, the correctness, and F, the fit).

      • hand_inundation_extent: The flood inundation extent based on HAND method.

      • observed_inundation_extent: The flood inudnation extent from the classified observed flood from Planet CubeSat satellite.

      • InterpolatedStage.csv: The estimated flood stages from the synthetic rating curves based on the flood discharge (observed or modeled). This is the result of script number 5.

  • Metric:
    Results of evaluation metric values (C, the correctness, and F, the fit) for the entire domain for all scenarios. In addition, for scenario 6, there is another csv file containing evaluation metric values for each catchment. There are three sub directories based on the Manning's n value that is used.

    • n005:

      • catchments: evaluation metrics values for each catchment as a csv file. Only available for scenario 6.
      • entire_domain: metrics values for the entire domain csv file
    • nOptimal: Only available for scenario 6.

      • catchments: evaluation metrics values for each catchment as a csv file.
      • entire_domain: metrics values for the entire domain csv file
    • n0018299: Only available for scenario 6.

      • catchments: evaluation metrics values for each catchment as a csv file.
      • entire_domain: metrics values for the entire domain csv file
  • Optimization:
    Results of the optimizaiton procedure (only available for scenario 6). First, using the script number 9, we compute the fit metric (F) for a range of stage values. The results are called, for example, "metric-for-stages-catch1.csv". Then, script number 10 is used to find the optimal stage value for each catchment based on the optimal (maximum) fit (F metric) value. The result is a csv file called "best-stage-basedonF.csv". Using script number 11, the optimal average inudnation depth is found (best-avginundepth-basedonF.csv). Finally, script number 12 computes the optimal local Manning's n. The result is called best-n-basedonF.csv.

Credits

Funding Agencies

This resource was created using funding from the following sources:
Agency Name Award Title Award Number
Utah Water Research Laboratory

How to Cite

Garousi-Nejad, I., D. Tarboton, M. Aboutalebi, A. F. Torres-Rua (2019). Data For Terrain Analysis Enhancements to the Height Above Nearest Drainage Flood Inundation Mapping Method, HydroShare, http://www.hydroshare.org/resource/665dfa6aee8d4689ab59def52b3b1179

This resource is shared under the Creative Commons Attribution CC BY.

 http://creativecommons.org/licenses/by/4.0/
CC-BY

Comments

There are currently no comments

New Comment

required