Writing prompt Grade 12 Science
Rennick Creek: Channel Change and Late-Summer Flow
A grade 12 scientific argument prompt in which students weigh channel incision against irrigation withdrawal as the cause of a fictional creek's summer flow loss.
Objectives
- Students will describe a feedback loop linking riparian clearing, channel incision, and floodplain water storage.
- Students will argue from a long-term data record for one causal explanation over a competing one.
- Students will estimate the size of a competing effect before deciding how much weight it deserves.
Vocabulary
- Baseflow
- Baseflow is the water that keeps a stream flowing even when it has not rained for a while. It comes from groundwater slowly seeping into the streambed.
- Water table
- The water table is the underground level below which the ground is completely soaked with water. Dig a hole deeper than the water table and it will start filling with water.
Also: riparian, channelization, incision, feedback
Teacher overview
Rennick Creek runs through the fictional township of Kestwick, where farmers cleared the streamside woodland and straightened the lower channel during the 1970s. Late-summer flow in the creek has fallen steadily since. Two explanations circulate locally, one blaming irrigation wells and one blaming the channel work itself, and the township's survey and gauge records contain enough information to weigh them. The prompt is built so that the explanation blaming withdrawal has to be tested against the size of the withdrawal actually permitted rather than assumed.
Prompt
Background. A creek in late summer is fed almost entirely by baseflow, which is groundwater seeping out of the banks and bed after surface runoff has ended. The floodplain beside a creek acts as a shallow reservoir. Water soaks into it during wet months and drains slowly toward the channel through the dry months. How much water that reservoir can hold depends on where the creek's bed sits relative to the floodplain surface. A creek that cuts down into its bed pulls the surrounding water table down with it.
Riparian woodland is the band of trees and shrubs along a stream. Roots hold the banks, shade lowers the water temperature, and fallen wood slows the current. When a channel is straightened, it is made shorter and steeper over the same drop in elevation, so water moves through it faster during storms.
The table below gives records kept by the fictional Kestwick township for the lower five miles of Rennick Creek. Benchmark K-3 is a fixed survey marker beside the channel, and well W-2 sits in the floodplain about 300 feet from the bank.
| Measure | 1972 | 1992 | 2012 | 2024 |
|---|---|---|---|---|
| Ten-year mean annual precipitation (inches) | 38 | 37 | 39 | 38 |
| Lower-creek bank length with continuous riparian woodland (miles, both banks combined) | 7.4 | 3.1 | 1.2 | 1.0 |
| Channel bed elevation at Benchmark K-3 (feet above datum) | 100.0 | 98.6 | 96.9 | 96.4 |
| August depth to water table at well W-2 (feet below ground) | 3.2 | 5.0 | 7.1 | 7.6 |
| Median August flow (cubic feet per second) | 14.0 | 9.5 | 4.8 | 4.1 |
| Hours from peak rainfall to peak creek flow | 11 | 7 | 4 | 4 |
| Irrigated acreage in the lower valley (acres) | 640 | 690 | 730 | 740 |
| Permitted summer irrigation withdrawal (acre-feet per season) | 210 | 240 | 265 | 270 |
One acre-foot equals 43,560 cubic feet, and the irrigation season in Kestwick runs about 90 days.
Writing task. Write a scientific argument for the Kestwick conservation district. Claim which of the two explanations, channel change or increased irrigation withdrawal, better accounts for the loss of median August flow since 1972. Support the claim with values from the table, describe the feedback loop your explanation depends on, and address the competing explanation directly rather than ignoring it. Estimate the flow that the increase in permitted withdrawal could account for, and use that estimate as evidence. State clearly what the record cannot settle.
Planning Organizer
Complete every row before drafting. The rows match the rubric criteria.
| Planning step | Your notes |
|---|---|
| My claim, in one sentence | |
| The total change in median August flow, 1972 to 2024 | |
| Step 1 of the feedback loop: what clearing and straightening did to storm velocity | |
| Step 2: what faster storm flow did to the channel bed | |
| Step 3: what a lower bed did to the floodplain water table | |
| Step 4: what a lower water table did to August baseflow | |
| Step 5: how the drier bank affects the remaining riparian trees, closing the loop | |
| Evidence rows I will cite for each step | |
| The competing explanation, stated fairly | |
| My estimate of the flow the added withdrawal could explain, with the arithmetic | |
| Why precipitation is not the explanation | |
| What the record cannot settle |
From Water for Schools (waterforschools.org). This material is in the public domain (CC0) and may be copied, edited, and shared freely.
Exemplar & Rubric
Exemplar response
The loss of late-summer flow in Rennick Creek is better explained by the channel work of the 1970s than by irrigation. Median August flow fell from 14.0 to 4.1 cubic feet per second between 1972 and 2024, a loss of 9.9 cubic feet per second. The record shows a self-reinforcing chain of channel and groundwater changes large enough to account for a loss of that size, while the growth in permitted withdrawal is far too small.
The chain begins with the clearing and straightening. Continuous riparian woodland along the lower creek's banks fell from 7.4 miles to 1.0 mile. A straightened channel is shorter and steeper over the same elevation drop, and a channel without roots, shade, and fallen wood offers less resistance. Storm water therefore moves through it faster, which the record captures directly: the interval from peak rainfall to peak creek flow shortened from 11 hours to 4. Faster water carries more sediment, and a channel that carries away more sediment than it receives cuts downward. Benchmark K-3 records exactly that, with the bed dropping 3.6 feet over the period.
An incised bed lowers the water surface the floodplain drains toward. Groundwater beside the creek flows toward the channel, so when the channel deepens, the surrounding water table follows it down. August depth to water at well W-2 increased from 3.2 to 7.6 feet, a drop of 4.4 feet that closely tracks the 3.6 feet of bed incision. The floodplain now holds less water above the level of the channel, and it is that stored water that seeps out as baseflow in August. Less stored water above channel level means less August flow, which is the observed result.
The loop closes at the bank. As the floodplain water table drops below the rooting depth of willows and cottonwoods, the remaining riparian trees die back. Their roots no longer hold the banks, erosion continues, and the channel keeps adjusting. The record is consistent with this: riparian length continued falling after the original clearing ended, from 3.1 miles in 1992 to 1.0 mile in 2024, and the bed kept dropping through the same period. A one-time disturbance produced continuing change, which is what distinguishes a feedback from a single event.
Irrigation deserves a fair hearing, because withdrawal did grow. Permitted summer withdrawal rose from 210 to 270 acre-feet, an increase of 60 acre-feet per season. Converting, 60 acre-feet is 60 times 43,560, or 2,613,600 cubic feet. Spread across a 90-day season, that is 2,613,600 divided by 7,776,000 seconds, or about 0.34 cubic feet per second. The increase in withdrawal can therefore account for roughly 0.34 of the 9.9 cubic feet per second that was lost, which is about one part in thirty. Even if every drop of the added withdrawal came straight out of August streamflow, it would leave nearly all of the decline unexplained.
Precipitation can also be set aside. The ten-year mean stayed between 37 and 39 inches across the whole record, with no trend, so the basin was not receiving less water.
Two limits are worth stating. First, the record is correlational. It shows bed incision, water-table decline, and flow loss moving together, but it contains no controlled comparison with an unaltered reach. Second, the withdrawal figures are permits rather than metered use, so actual pumping could exceed or fall short of them. A paired study on a nearby unstraightened creek, and metered pumping records, would strengthen the case considerably.
Rubric
| Criterion | Emerging | Developing | Proficient | Advanced |
|---|---|---|---|---|
| Claim and causal chain | States a preference with no mechanism | Names channel change as the cause but describes fewer than four steps | Describes at least four linked steps from clearing through incision and water-table decline to reduced baseflow | Describes the full loop, including how the drier bank feeds back on riparian cover, and explains why that makes the change self-sustaining |
| Use of the record | Cites no values, or values that do not bear on the claim | Cites values without tying them to a step in the chain | Ties at least four rows of the table to specific steps in the argument | Ties rows to steps and also uses the flat precipitation record to rule out a third explanation |
| Handling the competing explanation | Ignores irrigation entirely | Mentions irrigation without evaluating its size | Estimates the flow the added withdrawal could account for and compares it with the total loss | Carries out the conversion correctly, states the result as a fraction of the total loss, and concedes what irrigation could still contribute |
| Scientific writing and limits | Terms misused; no statement of limits | Terms mostly correct; limits stated vaguely | Riparian, incision, baseflow, and water table used correctly, with at least one stated limit of the record | Precise terminology throughout, with limits named specifically and paired with the additional evidence that would resolve them |
Differentiation
Support
Give students the five steps of the feedback loop on separate strips of paper in scrambled order and have them arrange the strips before writing, then write one sentence of evidence beneath each. Provide the withdrawal conversion partially worked, through the 2,613,600 cubic feet step, so students only complete the division by seconds.
Extension
Ask students to predict what the table would look like in 2044 if the township restored a continuous riparian buffer beginning next year, giving a direction of change for each row and a reason. Require them to identify which row would respond first and which would respond last.
ELL supports
Pre-teach riparian, channelization, incision, and baseflow with a labeled cross-section drawing showing the channel, the floodplain, and the water table before and after incision. Allow the Planning Organizer to be completed in phrases, and provide a causal sentence frame such as "Because the channel bed dropped ___ feet, the water table ___, which reduced ___."
Extensions
Have students locate a real stream-restoration project that raised a channel bed or reconnected a floodplain, then write a short technical note comparing the project's stated goals with the mechanism they described for Rennick Creek. Discuss why restoration projects usually target the channel geometry rather than the flow itself.