Geographical skills & fieldwork (incl. calculations)

The data largely support the hypothesis. Pedestrian numbers fall from 142 at the town centre (0 m) to 23 at 700 m, a decrease of about 84%¹. The mean count for the four sites closest to the centre (0–300 m) is about 114, compared with 46.5 for the four furthest sites, so the inner sites had more than twice as many pedestrians² on average.

However, the decrease is not consistent: the count rises at site 4 (300 m) to 101³ from 96 at site 3, and at site 7 (600 m) to 52 from 47 at site 6. Overall, the data support the hypothesis to a large extent⁴, but these two anomalies mean the pattern is not perfect.

Our hypothesis was that channel width, depth and velocity increase with distance downstream, as predicted by the Bradshaw model¹, because tributaries add more water and the channel becomes more efficient. We measured five sites along a small stream on one day in June.

Our results largely supported the hypothesis. Width increased from 1.2 m at site 1 to 4.6 m at site 5, an increase of about 283%². Mean depth rose from 0.11 m to 0.35 m, about 218%, and velocity increased from 0.21 m/s to 0.47 m/s, about 124%. As a result, discharge increased from 0.028 m³/s to 0.757 m³/s³, about 27 times larger. Depth increased at every site, so this part of the hypothesis was fully supported⁴. This fits the theory: tributaries joined between our sites, adding water, and the larger, deeper channel downstream has less friction relative to its volume, so water flows faster.

However, the results were not consistent at every site. At site 3, width fell slightly⁵ from 2.1 m to 2.0 m and velocity fell by about 15%, from 0.34 m/s to 0.29 m/s. This may be because site 3 was on a bend where a fallen branch and thick vegetation slowed the flow, and the bank was harder to measure accurately. So width and velocity only increased overall, not at every site.

The reliability of our conclusion is limited. We only sampled five sites⁶, so one unusual site had a big effect on the pattern. Velocity was measured by timing a float over 10 metres; although we repeated each measurement three times and took a mean, the float was sometimes caught in eddies or blown by the wind. We also collected data on a single day after a dry spell, so the results may not represent the river at higher flows.

Overall, our results supported the hypothesis to a large extent⁷, as all three variables and discharge increased clearly from site 1 to site 5. However, the anomaly at site 3 and the small number of sites mean we can only be partly confident, and more sites sampled at regular intervals on several dates would make the conclusion more reliable.