Leon Creek Watershed October 17-18, 1998 Rainfall Analysis Examination of USGS Gauge 8181400 Helotes Creek at Helotes, Texas Terrance Jackson MSCE Candidate University of Texas San Antonio Abstract The civil engineer often examines a storm event leading to flooding over a watershed. Gage stations are located on streams, lakes, or reservoirs where observations and hydrologic data are needed. The engineer performs quality control analysis of the radar rainfall data of the event and compares it to estimates from the rain gauge network. This can be accomplished by analyzing the precipitation grid and calculating the average rainfall in the study area and comparing it to the rainfall tables. The Leon Creek watershed was selected as the area of study. One hour raster are created to show the flooding areas in the watershed as the storm progress through the watershed. The mean rainfall (4.43 inches and 8.32 inches) for October 17-18, 1998 storm are equivalent to the 25 year storm. The maximum rainfall that occurred on day 2 was 12 inches and is equivalent to the 100 year flood. Introduction Leon Creek Watershed is located in northwest San Antonio (Figure 1). Civil engineers and hydrologists use rainfall data for watersheds to evaluate the rainfall distribution. The rainfall distribution changes with time as the storm progresses. NEXRAD total storm precipitation data is a series of maps that estimate the precipitation accumulation every hour. Polygon shapefiles can be downloaded and overlaid as a GIS layer. This research will show the progression of the storm to show flooded areas. These maps will be compared to the rainfall table (Figure 2) to show the impact of flooding in the Leon Creek Watershed. Data Used Rain gauge data was downloaded from USGS Real Time Water Data for the Nation. The website allows the user to choose the days of the storm event to retrieve the needed information for the gauge. Precipitation data was downloaded from National Oceanic and Atmospheric Data Administration (NOAA) website. This precipitation data is in real time or archive storm events can be downloaded. 1
Methodology: Rain gauge data for USGS rain gauge 8181400 is downloaded from the USGS website http://waterdata.usgs.gov/nwis/rt (Figure 3). Texas was selected as the Geographic area of interest. Daily Stage was selected as the predefined display. Group table by county was selected. Bexar County Station 08181400 Helotes Ck at Helotes Texas was selected. Available parameter 00065 gauge height (Max., Min., Mean) begin date October 17, 1998 and end date October 18, 1998 table format was selected (Figure 4). Precipitation data for the storm event was downloaded from the NOAA website http://www.ncdc.noaa.gov/nexradinv/chooseday.jsp (Figure 5). The Austin/San Antonio NEXRAD data covers a 230 km area which encompasses Bexar County (Figure 6). The National Climatic Data Center Inventory page requires user to choose date, product and select create graph to retrieve data. Order data is selected, the files are chosen, and an email address is entered. NCDC assigns the order a HAS number. The HAS number is used to retrieve the data. The analysis of the total storm precipitation image shows the precipitation data is measured in inches (Figure 7). NEXRAD data exporter is selected and vector (polygon) shapefile is chosen as the file format. Choosing the output directory is the final step to retrieve precipitation data. Eighty four shapefiles were exported for October 17 and the precipitation retrieval procedure was repeated to download the shapefiles for October 18. Test 1 folder was created for the rainfall data. Test 2 folder was created for the Leon Creek Watershed shapefile. Test 3 folder contains the Leon Tools box that shows functions used to create rainfall maps (Figure 8). A personal geodatabase (input rainfall.mdb) was created to store the precipitation (Figure 9). Precipitation data was imported to input rainfall.mdb using feature class to geodatabase (multiple) script (Figure 10). The rainfall polygons were in the GCS_WGS_1984 coordinate system. Therefore, the project tool was used to convert the watershed to the same geographic coordinate system (Figure 11). The clip feature script was used to clip the rainfall to watershed (Figure 12). Feature to raster (batch) tool converted the clipped rainfall polygons to raster file. The environment settings were changed to set the current workspace and scratch workspace (Figure 13). The mosaic to new raster tool was used to create one hour raster to study the rainfall distribution (Figure 14). 2
Results Figures 15-20 show the rainfall distribution in the Leon Creek Watershed for the storm event. Analysis of the precipitation data indicates that there is no precipitation recorded on October 18 from 5:00 am to 1:00 pm. However, the radar continued to accumulate precipitation on day 2. Figure 15 shows that a 10:00 am the1 inch of precipitation occurred in the northwest portion of the watershed. Figure 16 indicates that by 3 pm every pixel in the watershed received rainfall (1 to 6 inches). Figure 17 raster depicts that the area of concern are the 5 and 6 inch precipitation depths located in the center of the watershed. Figure 18 indicates that by 11:00 pm on day 2 the maximum precipitation increased to 8 inches. On October 18 at 3:00 am the 6 inch precipitation migrated to the southern part of the watershed. The rainfall raster at 11:00 pm indicates the precipitation total at the end of the event. 6 to 10 inches of precipitation encompasses most of the watershed. Conclusions and Discussion The rainfall distribution maps show that the entire watershed received rainfall during this two day storm event. The one hour rasters created show the rainfall distribution and can be compared to the rainfall table. The summary statistics of the first day of rain show that the maximum precipitation was 8 inches. The mean rainfall was 4.43 inches. The mean rainfall and maximum precipitation are equivalent to the 25 year event. The summary statistics of the second day of rainfall show a maximum precipitation of 12 inches. The maximum precipitation is equivalent to the 100 year event. The mean rainfall for the second day is 8.32 inches. The mean rainfall is equivalent to a 25 year event. References 1. USGS Real Time Water Data for the Nation http://waterdata.usgs.gov/nwis/rt 2. NOAA National Climatic Data Center Inventory http://www.ncdc.noaa.gov/nexradinv/chooseday.jsp 3
Figures 4
Figure 1 USGS Rain Gauge Locations Leon Creek Watershed USGS Rainfall Values (inches) Frequency of Storm 2- year 10- year 25- year 50- year 100- year 500- year Storm Duration 5 min 0.50 0.78 0.93 1.04 1.13 1.52 15 min 1.08 1.60 1.80 2.10 2.50 3.30 1 hour 2.00 2.76 3.32 3.85 4.35 5.80 2 hour 2.38 3.55 4.35 5.10 5.80 8.10 3 hour 3.00 3.95 4.90 5.70 6.60 9.40 6 hour 3.62 4.60 5.70 6.50 7.50 10.60 12 hour 4.15 5.40 6.40 7.50 8.80 12.40 24 hour 4.50 6.00 7.50 9.00 10.00 13.70 Figure 2 USGS Rainfall (inches) 5
Figure 3 USGS Gauge Data Figure 4 USGS gage (Maximum, Minimum, Mean) October 17, 18 1998 6
Figure 5 NEXRAD Data Search Page Figure 6 NEXRAD Data coverage 7
Figure 7 Total Storm Precipitation image Figure 8 Test 1, Test 2, and Test 3 folders 8
Figure 9 Create Personal Geodatabase for rainfall Figure 10 Feature Class to Geodatabase (multiple) script 9
Figure 11 Project watershed to GCS_WCS 1984 coordinate system Figure 12 Clip Features Script 10
Figure 13 Feature to raster (batch) and Environment settings Figure 14 One Hour Raster personal geodatabase 11
Figure 15 One hour raster October 17 10:00 am Figure 16 One hour raster October 17, 3:00 pm 12
Figure 17 One hour raster October 17 at 2100 hrs Figure 18 One hour raster October 17 at 2300 hrs 13
Figure 19 One Hour Rainfall Raster October 17 3:00 am Figure 20 One Hour Rainfall Raster October 18 11:00 pm 14
Figure 21 October 17 at 2351 hrs Statistics Figure 22 October 18 at 2355 hrs Statistics 15