## Info

Particle size (um)

FIGURE 3.4 Inter-storm variation in the EPSD on the Upper Exe at Pixton, and the Exe at Thorverton.

of the full EPSD for each periodic measurement), together with the mean EPSD for a summer event on the River Exe (no summer sample available for the Upper Exe). The greater difference between the summer and winter events on the River Exe, relative to the two consecutive winter events, clearly necessitates a consideration of the relative importance of inter-storm variations in the EPSD and DOA over seasonal time scales.

Figure 3.5 presents the storm-period mean values of Ed50 and the DOA for all study sites plotted against the Julian day (day 1 being January 1) of the storm event. A clear seasonal trend is evident in both data sets, with the Ed50 and DOA being at a minimum during the mid-winter period of December and January (days 335-331), increasing through the spring to a summer-autumn (May to October; days 121-304) maxima before a general reduction in both parameters is evident toward winter.

The statistical significance of the observed seasonality in the EPSD and DOA of the study sites was assessed by Mann-Whitney U-test analysis of the summer/autumn and winter/spring values of mean and median effective particle size, as well as the degree of sorting, and the DOA. Analysis was restricted to those sites considered to have a sufficient seasonal range in samples (all except the River Clyst, Upper Exe, and River Creedy, with data from the Jackmoor Brook and Yendacott being amalgamated to be representative of the Jackmoor Brook as a whole). This analysis established that for both the EPSD and DOA there was a significant difference (0.05 level) in all three variables between the summer/autumn and winter/spring grouped data sets. The lack of a statistically significant seasonal difference in the three parameters for the seasonally grouped APSD data emphasizes the importance of seasonal variability in aggregation in influencing the patterns described above.

Julian day

FIGURE 3.5 Seasonality in the magnitude of the storm-period mean values of Ed50 and the DOA at all sites.

Julian day

FIGURE 3.5 Seasonality in the magnitude of the storm-period mean values of Ed50 and the DOA at all sites.

The scatter in the relationships depicted in Figure 3.5 reflects in part spatial variation in the EPSD dynamics of the different rivers in the study area. Figure 3.6(a) and (b) present data for the study site at Thorverton on the River Exe only, this being the study site with the most complete seasonal data set and from which seasonality in the EPSD and DOA can best be considered without the additional influence of spatial variations in the data. In both Figure 3.6(a) and (b) the strength of the seasonal trend in the storm event mean value of both Ed5o and DOA, as described by the value of r2 for the polynomial relationship, is higher than that determined for the study area as a whole in Figure 3.5. Similarly strong seasonal trends are evident for the maximum and minimum storm event values of Ed5o and DOA that are also shown in

FIGURE 3.6 (a) Seasonality in the magnitude of the storm-period mean, minimum and maximum value of Ed5o for the River Exe at Thorverton and (b) seasonality in the magnitude of the storm-period mean, minimum and maximum value of the DOA for the River Exe at Thorverton.

### Julian day

FIGURE 3.6 (a) Seasonality in the magnitude of the storm-period mean, minimum and maximum value of Ed5o for the River Exe at Thorverton and (b) seasonality in the magnitude of the storm-period mean, minimum and maximum value of the DOA for the River Exe at Thorverton.

Figure 3.6(a) and (b). These data, together with those of Figure 3.5, demonstrate that seasonal variations in the EPSD and DOA are significant phenomena, and add a further temporal dimension to any attempt to characterize the hydrodynamic properties of fluvial suspended sediment in the study area.

### 3.4.5 Effective Particle Density

Temporal variations in the DOA have been demonstrated to be the major determinant of the EPSD of fluvial suspended sediment in the study area. The significance of aggregation upon the second key hydrodynamic property of fluvial suspended sediment in the study area, namely, particle density is considered below.

Figure 3.7 presents the results of the five quantitative assessments of EPD undertaken using the Par-Tec 200 laser backscatter probe and the Valeport (SK110) BWT. These data indicate that the median EPD of all samples is considerably less than the ca. 2.65 gcm3 that could be expected from a dispersed sample of mineral sediment with the same size distribution. Furthermore, an inverse size/density relationship is apparent, which is in accordance with that previously reported for fluvial suspended sediment.11'18'21

When considering the significance of the EPD data presented in Figure 3.7 to the intra-storm and seasonal variations in effective particle size described above, it should be noted that the relationship between effective particle size and density presented in Figure 3.7 is generalized both spatially and temporally. With this caveat in mind, it is reasonable to conclude that EPD will vary significantly throughout the study area, both at the intra-storm and seasonal temporal scales. For example, application of the power function contained in Figure 3.7 to the Ed50 data in Table 3.2 results in an intra-storm range in EPD of between 0.01 gcm3 (Upper Exe, 19/12/93,20/12/93, and 03/02/94 and Exe 13/09/93) and 0.35 g cm3 (Jackmoor Brook 08/04/94), with a study area average intra-storm range of 0.10 gcm3. Seasonal variation in EPD is illustrated in Figure 3.8 which presents the estimated mean storm-period EPD using the power

FIGURE 3.7 The inverse relationship between median effective particle size and median particle density determined for the study area.

Ed5o (urn)

FIGURE 3.7 The inverse relationship between median effective particle size and median particle density determined for the study area.

Julian day

FIGURE 3.8 Estimated seasonal variation in effective particle density for the combined study site data set.

Julian day

FIGURE 3.8 Estimated seasonal variation in effective particle density for the combined study site data set.

function contained in Figure 3.7. As would be expected, the trend approximates the inverse of the seasonal trend in EPS, with EPD being a minimum during the summer-autumn period.

## Post a comment