Sampling within a drainage basin
Even if you are just focussing on a short stretch of a stream, it is not necessary to measure every part of the river. Instead you need to design a sampling strategy that will yield representative data without bias.
For investigating downstream changes in one river, the most straightforward way to do this is to take a systematic sample. Select 1015 sample points at regular intervals (e.g. a 50m distance between each site). Your choice for the distance between sample points will need to be justified; make sure that they are representative of the whole length of the river. You may also wish to make sure that there are sample points just upstream and just downstream of major features of interest, such as a point where tributaries join or where water is abstracted. If you are comparing the characteristics of more than one stream, make sure that you that use the same sampling strategy for each.
Measuring river variables
Depth, width and wetted perimeter  Velocity and discharge  Long profile  Bedload
(i) Depth, width and wetted perimeter
At each sample site, measure
(a) Channel width  use a tape measure or rope held across the channel from bank to bank on the water surface.
(b) Channel depth (or wet depth)  take a number of readings across the channel (e.g. ¼ across,1/2 across, ¾ across) to measure the depth from the water surface to the stream bed.
(c) Wetted perimeter  use a tape measure or chain to measure the total distance that water is in contact with the bed and the banks. You may need to secure the tape in position with stones or pegs, especially if the water is fastmoving.
The channel width, channel depth and wetted perimeter vary from day to day. If you are are sampling on one day only this is not a problem, but you can obtain additional information by measuring the bankfull width, depth and wetted perimeter as well. Bankfull is the stage where the river is completely filling its channel.
At each sample site also measure
(d) Bankfull width  use a tape measure or rope held across the channel from bank to bank. The edge of the channel may be marked by a change of slope and/or vegetation. Keep this in position to help you with (e).
(e) Dry depth  measure the depth from the bankfull height to the water surface. If you take readings in the same place as you measured channel depth (or wet depth), you can add the two figures together to find bankfull depth.
(f) Bankfull wetted perimeter  use a tape measure or chain to measure the distance from bankfull point on one bank to bankfull point on the other.
(ii) Velocity and discharge
The cheapest way that you can measure velocity and discharge is the float method. You need a 10m tape measure, a float (i.e. something that floats and is brightly coloured, such as an orange) and a stopwatch. You also need at least two people for this method.
Measure a set distance of the river and mark the start and end points. 10 metres is a good idea  it's long enough to show interesting variation in results, and a round number also makes the maths much easier! Put the float in the water slightly upstream of the start point. Using the stopwatch, time how long it takes the float to move from the start point to the end point. Repeat this procedure at least five times, placing the float at regular intervals across the stream so that you measure velocity across the channel.
It's pretty likely that at some stage the float will get stuck in an eddy! Either nudge the float to move it along or abandon that reading and start again.
The raw data that you have collected by this technique can be used to measure velocity and discharge. See Data Presentation.
(iii) Long profile
Although you can plot long profiles using a good 1:50,000 topographic map, for more detailed information you should collect data in the field. You need a 10m tape measure, two ranging poles and a smart phone adapted for use as a clinometer. You also need at least two people for this method.
To avoid standing in the river, you can use the gradient of the dry land on the banks as a substitute for the water surface slope. At each sample point on the river's course, measure slope angle from the sample site upstream and slope angle from the sample site downstream. See the diagram for how to do this.
(iv) Bedload
A river's load can be classified into three groups: bedload (sand, pebbles and stones), suspended load (clay and silt) and dissolved load (soluble material). Although all types of load can be measured in the field, in practice measuring the bedload is the most stratightforward of the three.
Since both the volume of load (measured in cubic metres) and the competence of the river (the largest sized particle that the river can transport) increase exponentially when the river is in times of flood, samples of supended load made at times of low flow will not be representative of the whole year's load.
It is not, of course, possible to measure every stone in the river. Instead you will need to design a sampling strategy to give you a representative sample without bias that you can repeat at each sampling point. For example, you could collect 20 pebbles at random from the river bed at three different points across the width of the channel.
The simplest way to measure pebble shape is to classify the stone as very angular, angular, subangular, subrounded, rounded or very rounded. Decide which shape is the best fit for each pebble.
very angular
For an estimate of pebble size, measure the longest axis of each pebble.
For pebbles in the rounded category, a numerical value of degree of roundness can be calculated. The raw data needed for each pebble is as follows.

The length of the longest axis (called a)

The length of the vertical axis (b)

The length of the horizontal axis (c)
b + c / 2a x 100 gives an index of roundness in %
This will also give you more precise size data. For pebbles where it is difficult to pick out the axes, allow the pebble to rest on a flat surface. The length of the longest axis is the a axis.
angular
subangular
subrounded
rounded
very rounded