4.1 Overview of the measurements

The idea behind the full-scale comparison test of the driving rain gauges is to register the readings of the different driving rain gauges while they are exposed to the same, natural driving rain for a reasonable period of time. In order to expose the gauges to the same driving rain, they were mounted together as closely as possible on the west façade of the Main Building of the TUE (positions P4, P5 and P7; P4/5/7 for short). Figure 3.8 shows a photograph of the situation. The spatial positions of the gauges are within 2.5 m horizontally and within 0.7 m vertically. The exact operation periods and the exact positions of the mounted driving rain gauges are listed in table 3.3 and described in section 3.2. The periods during which the gauges were operational, are also visible at a glance in table 4.1, in which monthly driving rain amounts of each gauge at P4/5/7 are listed. This table also lists the percentage of clock periods with available data (i.e. when the installed devices worked properly).

Table 4.1: Monthly driving rain amounts [mm] at position P4/5/7 of four different driving rain gauges: TUE-I (and TUEIb with a reduced catchment area), TUE-II, CTH and DTU. The last column shows the percentage of clock periods with available data. An asterisk indicates that the device was not operational during the whole month, but (un)installed during the month.

	TUE-I	TUE-Ib	TUE-II	CTH	DTU	% avail.
1997 12	1.06	-	2.78	-	-	99
1998 01	2.88	-	5.77	-	-	97
1998 02	0.49	-	0.83	-	-	98
1998 03	2.06	-	4.87	-	-	82
1998 04	0.29	-	1.11	-	-	79
1998 05	0.37	-	0.94	-	-	98
1998 06	1.52	-	3.92	-	-	70
1998 07	$\ast$ 0.00	-	1.93	$\ast$ 1.69	$\ast$ 1.80	90
1998 08	-	-	3.35	4.31	4.47	99
1998 09	$\ast$ 0.17	-	9.61	10.93	9.37	99
1998 10	6.40	-	11.70	10.93	8.96	88
1998 11	0.12	-	0.59	0.81	1.19	82
1998 12	1.88	-	3.43	3.40	3.78	99
1999 01	1.38	-	3.30	2.76	4.68	99
1999 02	$\ast$ 1.95	$\ast$ 0.00	4.06	1.59	4.06	98
1999 03	-	$\ast$ 0.97	4.31	0.45	5.07	99
1999 04	-	0.27	1.49	1.10	2.22	98
1999 05	-	0.01	0.12	0.29	0.56	99
1999 06	-	0.98	2.43	2.66	3.75	99
1999 07	-	0.77	2.25	1.62	3.47	90
1999 08	-	1.08	1.70	1.20	2.88	89
1999 09	-	0.12	0.76	0.52	1.45	96
1999 10	-	0.14	0.82	0.19	1.30	99
1999 11	-	0.04	0.80	0.26	1.39	94

The TUE-II gauge will often be used as a reference for the comparisons. It is the gauge which was operational for the whole 24-month measurement period. Moreover, as we will see in the following, it is one of the gauges which functions well. As monthly driving rain amounts do not give many conclusive details on the performance of the different gauges, we mention the monthly results of a gauge compared to the TUE-II gauge only concisely here:

• CTH $\leftrightarrow$ TUE-II. The CTH gauge functioned from 17 July 1998 to 31 November 1999. For this period, its total driving rain amount was 85% of that of the TUE-II gauge.
• DTU $\leftrightarrow$ TUE-II. The DTU gauge functioned during the same period as the CTH gauge. The total driving rain amount of the DTU gauge during that period was 116% of that of the TUE-II gauge.
• TUE-I $\leftrightarrow$ TUE-II. From 1-12-1997 to 7-7-1998 and from 16-9-1998 to 25-2-1999, the driving rain gauges TUE-I and TUE-II were both operational. They were mounted close to each other at façade positions P4 and P5; after 16 September 1998 the gauges were interchanged. Evidently, the TUE-II gauge registered more driving rain than the TUE-I gauge. The total driving rain amount of the TUE-I gauge during the mentioned period was 47% of that of the TUE-II gauge.
• TUE-Ib $\leftrightarrow$ TUE-II. In February 1999 the TUE-I gauge was adapted, i.e. its catchment area was reduced. It was (fully) operational from April to 31 November 1999. Its total driving rain amount was 32% of that of the TUE-II gauge.

In [Högberg et al. 1999], the readings of the four gauges CTH, DTU, TUE-I and TUE-II were compared for the period of 1-10-1998 to 28-2-1999. For these five months, the CTH, DTU and TUE-I gauges, respectively, measured 94%, 84% and 51% of the driving rain amount registered by the TUE-II gauge. Compared to the TUE-II gauge the DTU gauge measured less during this five-month period (84%) than during the 16-month period (115%). Also the CTH gauge seems to have a different performance after February 1999. It is impossible to check whether the performance of the gauges really changed over time. In the following two sections (sections 4.2 and 4.3) we will therefore compare the driving rain registrations on a smaller time basis, namely 10-min clock periods. On these results our explanation of the differences will be based. In general, differences in reading arise from:

a.

aerodynamic disturbance, by which raindrop trajectories deviate from those without the gauge. The cause may be protruding rims or other projections, or even a collector sunk into the wall,

b.

run-off water flowing from other parts of the façade into the collector,

c.

raindrops splashing out of the collector,

d.

size and shape of the catchment area,

e.

hydrophobicity and smoothness of the collector. This is particularly important if the collected raindrops have to find their way to the water flux gauge. This aspect is also related to whether and how raindrops coagulate before rolling downwards,

f.

drainage path. The length and `straightness' of the path which drops have to follow from impingement to the point where they are measured,

g.

influence of wind on the path of the drops. In extreme cases the wind may blow drops out of the collector, or prevent drops from getting into the water flux gauge,

h.

influence of wind on the reading of the DTU gauge (its collector is suspended freely from a strain gauge),

i.

evaporation of raindrops or collected rain water. Depending on the finish of the collector surface and the actual number of drops (raindrop spectrum and rain intensity), smaller drops remain stuck on the collector surface and are not measured,

j.

principle of the rain flux measurement. Important is its resolution, i.e. the minimum measurable amount of water during a clock period,

k.

spatial differences, because the gauges are not exactly mounted on the same position.

Aspects a, b and c were not investigated during the comparison test, although at designing and mounting of the gauges these aspects were taken into account. Aspects e, f and g were also taken into account at the design, whereas during the test, cleanness of the collector surface and the motion of drops through the drainage system of the gauges were inspected visually whenever possible. Differences due to aspects d, e and h-k are investigated by correlating the readings of the different gauges to each other, and by relating them with the reference rain intensity ( $R_{\text {h,c}}$) and wind speed ($U_y$). Of course, the raindrop spectrum is an important factor for all the mentioned aspects where rain intensity plays a role. Raindrop spectra were measured from October 1999 to January 2000. However, they are not included in this chapter, because during this period the amount of driving rain was unfortunately unsufficient for a meaningful correlation between raindrop spectrum data and driving rain data.

Figure 4.1: Correlations of 10-min driving rain intensities $R_{\text {f}}$ of gauge CTH, DTU and TUE-I with gauge TUE-II, for selected reference wind velocity components perpendicular to the façade $U_y$ of 4-5 m s$^{-1}$ (measurement points $\includegraphics[width=0.5em]{gen/m-kruisje.eps}$ and correlation $\includegraphics[width=2em]{gen/m-ononderbroken.eps}$) and 6-7 m s$^{-1}$ ( $\includegraphics[width=0.5em]{gen/m-rondje.eps}$ and $\includegraphics[width=2em]{gen/m-streep.eps}$). Period: 1-12-1997 to 30-11-1999.

Figure 4.2: As figure 4.1, but with extended axes.

Figure 4.3: As figure 4.1, but for reference rain intensities $R_{\text {h,c,\texttt {P2}}}$ of 1.0-2.5 mm h$^{-1}$.

Figure 4.4: As figure 4.1, but for reference rain intensities $R_{\text {h,c,\texttt {P2}}}$ of 3.0-5.5 mm h$^{-1}$.