Novato Creek Watershed

From Wildsong
Jump to navigationJump to search

2015 Aug 19 Greetings, professional hydrologist. Cut me some slack, this is my first attempt. You probably studied hydrology for 4 years. I've been at it for 2 days now.

I have spent the last few months exploring the hills in the Novato area, hiking and testing radios and geocaching.

Yesterday I started reading a little hydrology book. (Water, Rivers and Creeks, Leopold 1997) Leopold says some guy in 1600's noticed the amount of water flowing into the sea from the Seine was about the same as the rainfall. How did he get his numbers? I figure it must have been a pretty rough estimate.

Can I use modern tools to do my own estimate?

Oh heck, I should be looking at the Army Corps of Engineers Hydrologic Engineering Center models. In particular there is an ArcGIS extension, HEC-GeoHMS.


Stream Gage

Given that

  1. Input: a watershed basin contains the rainfall.
  2. Output: river(s) are outflow and soils absorb rainfall at different rates

Inputs:

  1. Find the area of the watershed.
  2. Find the annual rainfall for the area.

Outputs:

  1. Find the flow rates and volumes for the natural features.
  2. Find volumes for man-made features (outfalls).
  3. Calculate total volume.


The process:

  1. Pick an area. I need good data. Local water agency + USGS stream data + GIS.
  2. Build a model.
  3. Run the model.
  4. Look at results
  5. Refine the model.
  6. Go to step 3.

Area: Since I have spent so much time hiking and geocaching here, let's try the Novato Creek watershed

From Marin Watersheds: "The Novato Creek, located at the northwestern extent of San Pablo Bay, is the largest watershed in eastern Marin County. Its creeks flow eastward through oak and bay forests, grasslands, the City of Novato, and into San Pablo Bay near the mouth of the Petaluma River. The basin is 45 square miles and the main drainage in the watershed is Novato Creek; Novato Creek joined by six major tributaries along its 17 mile length: Leveroni, Bowman Canyon, Warner, Arroyo Avichi, Arroyo de San Jose, and Simmonds Slough."

Watersheds and flow

For process, refer to http://web.ics.purdue.edu/~vmerwade/education/hydrology.pdf

I created a model to do these steps. Data in lowercase, tools in Mixed_case.

Notes:

  • I need a shoreline polygon so I don't include the bay in the basin generation. That creates a zillion little polygons.

Find Watersheds

  1. Define ROI (region of interest) by drawing a polygon in ArcMAP -> roi
  2. n38w123_dem, n39w123_dem -> Mosaic_dataset ->
  3. Project_raster (into NAD_1983_HARN_StatePlane_California_III_FIPS_0403_Feet)
  4. Clip_raster using roi
  5. Fill -> // This gets rid of small basins that are just noise
  6. Flow_direction -> flow_dir ->
  7. Basin ->
  8. Raster_to_polygon -> watersheds

Find creeks and streams

  1. flow_dir -> Flow accumulation // shows where flow accumulates ie creeks and streams
  2. Raster_calculator // find areas with flow
  3. Raster_to_polyline -> flowlines
  4. Reclassify // for example flowacc<500 = intermittent flowacc>=500 = perrenial

I set the Raster_calculator to generate a layer in which output cells are either NODATA or the accumulated flow. That makes it easy to see what flow is at a given point. The expression I used is SetNull("flowacc"<100,"flowacc") Had to also do one in which the output was set to '1' for the Raster_to_polyline tool to work, another ESRI bug.

You could base the Raster_calculator threshold according to the area and the number of pixels in the area. That's what the above PDF tells you to do. 25 km2 = 25,000,000 meters2 / (30*30) = 27778... but I used 1000. (Trial and error method.) This will show ditches that might not even be classed as intermittent streams.

Data sources

Projections:

  • DEM is in GCS_North_America_1983 (unprojected)
  • Most of the county data is in NAD_1983_HARN_StatePlane_California_III_FIPS_0403_Feet

DEM - N39W123 NED data covers the area I am interested in. From USGS 3DEP. This is 1/3 arc second which is 10 meter pixels, okay for first pass testing of the model. Then I will want to get 1/9 arc second which is 3 meter pixels.

External links for resources; see also my Data sources page.

Novato Creek reports: http://marinwatersheds.org/novato.html

Local: Marin Watersheds Novato Creek reports

MarinMap where I got these: groundwater, nhd_flowline, waterbody, watershed_major, wetland

National:

NOAA Novato stream gage

USGS NWIS mapper | Novato Creek stream gage data

H&H study (Hydrology and hydraulics) was done by Kamman Hydrology and Engineering, Inc. in San Rafael