|
STAFF PAPER |
CSM, or cubic feet per second per square mile, is a demonstrated, useful unit of measurement in hydrology. It makes possible several techniques for increasing one's hydrologic awareness. It permits an assessment of overall river conditions at a glance, it gives a regional hydrologic awareness, it can help identify bad data points, and it allows estimation of flow in an ungaged basin or at a non-telemetered gage. It is often used in agencies outside the National Weather Service, and engineering hydrology studies are sometimes couched in terms of CSM's.
Let's look at CSM through a fictitious case. Suppose two river gages at Rock Creek and Falls River are reporting stages (that is, river levels) as follows:
| Rock Creek | Falls River |
| Stage = 14.9 feet | Stage=10.2 feet |
What can be said qualitatively about the flow state at each gage? The numbers should mean nothing to you, because they are fictitious; but in a real case they might also mean nothing unless you had memorized the whole range of significance of stages at these two gages. What else would you need to know to understand the level of flooding at each gage? Suppose you knew the flow in cubic feet per second (CFS) at each river at the given stage:
| Rock Creek | Falls River |
| Stage = 14.9 feet | Stage=10.2 feet |
| 2400 CFS | 10,000 CFS |
Does that help? Probably not, because large river basins tend to have large flows and small ones collect less water and generally have small flows. Since size of the drainage area has such an effect, we could assume drainage areas of 100 and 5000 square miles and then compare the flows to the respective areas.
| Rock Creek | Falls River |
| Stage = 14.9 feet | Stage=10.2 feet |
| 2400 CFS | 10,000 CFS |
| 100 Sq. Mi. | 5000 Sq. Mi. |
Comparing the ratio of flows to drainage areas:
|
|
The two ratios, 24 and 2, compare two flows to their respective drainage areas, and they are an indicator of the flow conditions at the two sites. These ratios are called "CSM values" and have units of cubic feet per second per square mile.
In the case above, the Rock Creek gage has a much higher CSM value than Falls River. In other words, the normalized flow at Rock Creek is much larger than that of Falls River. Below we will see further how to interpret CSM's. Rock Creek's CSM could represent a flood, while Falls River's is very ordinary.
[Top]
Employing CSM's is not an exact science, but still CSM's are useful. They must be tempered with knowledge of other factors and conditions, such as size of drainage basin, topography, and overall storage capability of soils -- all readily discernible information. A view of several actual cases should help. Table 1 is a list of CSM values and highly qualitative basin characteristics for 76 selected river gaging sites in the Northeast. These CSM values have been selected with something in common: they have been calculated for the flows experienced when these rivers are at flood stage (i.e., the river level where the water just begins to go over banks and cause damage.) It's as if we are imagining that all the rivers in the Northeast are bankfull. With that assumption, what can be generalized about the factors listed in Table 1 in regard to CSM's? It is clear from the table that CSM values tend to be higher in basins that are small, hilly, urban, or rocky. Moving down to the bottom of the table, it is also clear that CSM values tend to be lower in large, flat, or spongy basins. Experience has shown that this tendency also holds at uniform wetness regimes other than flooding. Figure 1 shows this general tendency of CSM's with regard to qualitative indicators of size, usage, topography, and porosity.
[Top]
4. Using CSM to Assess Overall River Conditions
Some actual stream gage data with CSM's generated by a computer is shown in Table 2. This data is typical of flow conditions several days or weeks after the last significant storms and floods. Rivers have receded almost down to base flow in most areas, with some notable exceptions. On May 12, 1997, the driest area, western New York, has the most uniform CSM's: Black Creek at Churchville, Genessee River at Avon, Oatka Creek at Garbutt, and Cayuga Creek at Lancaster have CSM's of 0.9, 0.7, 1.0, and 1.0, respectively. A moderately wet area in eastern New York and western Vermont is seen in readings from the Winooski River at Center Rutland (2.9), Passumpsic River at Passumpsic (3.9), Battenkill at Battenville (3.8), and Hudson River at Fort Edward (4.3). A wetter area in northern Maine and New Hampshire where snowmelt was still occurring is exemplified by the St. John River at Fort Kent (8.4), Penobscot River at West Enfield (5.2) , and the Pemigewasset River at Woodstock (4.8). Generally, under nearly steady state conditions and similar influences, large areas' flows will settle toward similar CSM values.
[Top]
Bad or anomalous data can be identified by scanning lists of CSM's from a region where flow conditions are known to be fairly uniform. Here is a recent case, depicted in Figure_2. On a particular day, CSM's for the Genessee River Basin were less than 1.0 for three out of four gages. However, the Rochester value was 2.8! It was known that there were no sources of water between Avon and Rochester that could have contributed such a large flow; therefore it was concluded that the Rochester CSM must be in error. On investigation it was found that the stage being reported was correct, but an error in a computer table led to the calculation of a flow of 7000 CFS, when it should have been 3000 CFS. The corrected CSM for this 2500 square-mile basin was a more plausible 1.2.
[Top]
6. Estimation of Flow by Transposition from nearby Basins
The tendency of neighboring basins to have similar CSM values under steady state conditions leads to another use of CSM's: flow may be estimated on ungaged rivers or at gages where there is no telemetry. Estimation of flow by extending or adopting CSM's in neighboring basins depends on knowledge of drainage areas. It also depends on the absence of perturbations or abnormal or human-controlled events, such as regulation due to power generation, strong local rainstorms, dam failures or other sudden water releases.
Figure 3 shows a hypothetical example of three neighboring basins, A, B, and C. All three drainage areas are known, as well as two of the basins' CSM values. If we assume uniform hydrologic conditions over all three basins, then we can assume the CSM in basin A is approximately 2; therefore its flow at Point A will probably be about 200 CFS (that is, 2 x 110, rounded to one significant digit.)
[Top]
What is the long-term mean CSM at the gage on the Connecticut River at Thompsonville CT?
What was the peak CSM during the April 1997 flood on the Red River of the North at Grand Forks, ND?
(By comparison, the flood of record at Hartford CT in 1936
was 313,000 CFS from 10,487 Sq.Mi. for 29.8 CSM.)
Congratulations! You have completed the CSM's module.
