Home  Newsletter Index    WMC   < Previous  TOC  Next >

Winter 1992

Low-Order Stream Channel Mapping
A new component of Coordinated Resource Inventory

Michael J. Furniss and Sam A. Flanagan
USDA Forest Service, Six Rivers National Forest


Figures are available in paper copy only.

Low-order stream channels in northern California forested watersheds are a special management problem. They are numerous, high gradient with steep sideslopes, difficult and dangerous to traverse, and constitute the most sensitive lands in all watersheds. Hillslopes adjacent to most low-order streams are typically steep, deliver eroded materials directly to channels, and have landslide-prone "valley inner gorge" morphology (Figures 1 and 2).

Knowing the extent and characteristics of the low-order streams in a planning watershed is critically important information - but is difficult to obtain and therefore usually missing. When such information is gathered, it has commonly been localized (e.g., the channel directly below a cut unit or at a road crossing), and a watershed-wide picture cannot be drawn.

As an part of our Coordinated Resource Inventory (CRI) - which includes the mapping of soils, geology, and vegetation, all at once - we are trying to develop practical techniques for mapping and describing all stream channels in a survey area. The large streams are easy - they are not extensive and are often described in fish habitat surveys. The small streams are the big challenge.

We had two primary questions:

"What techniques for stream channel and streamside slope mapping can produce a useful picture of the most sensitive lands in a planning watershed?"

"What does it take, in time and money, to complete a useful inventory of all stream channels and streamside slopes in the steep, dissected forested watersheds common in northern California?"

We assembled an interdisciplinary group and brainstormed what characteristics we were interested in. We then pared the list down to what seemed to be a practical minimum set of characteristics to measure. We then set out to survey all the streams in a 15,000-acre survey area, in the Bluff Creek watershed near Orleans California, which is tributary to the Klamath River. Streams were defined as:"any channel or declivity showing evidence of annual scour or deposition".

The survey technique divides the stream into short, successive sections based on geomorphic and riparian characteristics. Thus each section contains relatively consistent features. Mappable features within each section are:

A new section was begun when a "significant" change occurred in one or more of the above parameters. Of particular interest were the points of abrupt change along the stream channel. Such changes typically were associated with hillslope failures, tributaries, or sharp bends. For example, where a recent landslide had altered inner gorge morphology and riparian zone width. What constituted a significant change, on which to break the section, was the principal focus during the first several field trials. At first, delineation of sections was based on observer subjectivity. This encouraged valuable discussions among the crew as we considered whether a "change" had occurred. Our goal was to balance efficiency with mapping resolution. After several weeks of trial and error, the following characteristics were chosen to represent mappable changes (i.e., a new section was begun): Using these values, the mean section length for first and second order streams was 90 feet. Mean section length increased to 125 feet for third-order channels. Section length varied between 7 and 465 feet, with the longest sections found in the largest channels. Sections less than 25 feet typically were bedrock cascades.

No map is prepared in the field, but aerial photos are carefully referenced and control points are noted in the data. A schematic of a map can be prepared from the data (Figure 3). Automated map generation from a spreadsheet of the data is under development.

Data analysis and integration with geology, soils, and vegetation information, and application of GIS is underway. Many of the channels mapped are difficult or impossible to discern on aerial photos. Therefore, the technique can provide detailed information concerning stream protection and sensitive areas that is otherwise unobtainable. As a first-cut, we believe that this technique is practical, but needs to be refined to reduce the amount and/or resolution of data taken to increase production and decrease costs.

Production

A crew of two very fit earth scientists was able to map 2600 feet of channel length per day, equivalent to a 100-acre watershed. Production was somewhat proportional to the size of the streams. Overall costs were about $2.40/acre in this terrain where drainage density is especially high. This figure might be surprisingly high, but the need for and utility of the information is also high. The authors welcome all questions and comments.

You can contact Sam and Mike at (707) 441-3551 FTS 448-3551


Top