4.1 A NEW HIGH ACCURACY, LOW MAINTENANCE ALL WEATHER PRECIPITATION GAUGEFOR METEOROLOGICAL, HYDROLOGICAL AND CLIMATOLOGICAL APPLICATIONS

Heikki Turtiainen *, Pauli Nylander and Pekka PuuraVaisala Oyj, Helsinki, Finland

Risto HölttäVaisala Inc, Boulder, Colorado

1. INTRODUCTION

Accurate measurement of precipitation isa challenge, especially in climatic conditionswhere both liquid and solid precipitation occur.In principle weighing gauges are the mostsuitable point precipitation gauges for theseconditions. However, the accuracy ofconventional weighing gauges is degraded by amultitude of errors. Wind effects, evaporationand wetting error tend to cause systematicunderestimation. In winter conditionsinstrumental errors related to accumulation ofsnow and ice on rim and funnel parts of thegauge - including complete filling of the gaugewith snow - may result in gross errors. Theseproblems are only partially solved by usingantifreeze solution in the container and with rimheating.

The new Vaisala all weather precipitationgauge VRG101 (fig. 1) is a result of carefulstudy of the existing solutions and newtechnical innovations. It provides higher qualityof measurement and lower life-cycle cost in allweather conditions.

This paper describes the gauge designand presents field test results obtained at theobservatories of the Finnish MeteorologicalInstitute in Jokioinen and in Sodankylä.Preliminary results from field tests by NADP(National Atmospheric Deposition Program) arealso discussed.

* Corresponding author address: Heikki Turtiainen,Vaisala Oyj, PB 26, FIN-00421 Helsinki, Finland;e-mail: heikki.turtiainen(at)vaisala.com.

Fig. 1. Vaisala all weather precipitation gaugeVRG101 with wind shield VRS111.

2 VRG101 DESIGN

2.1 Design principles

VRG101 is designed as a reliable andaccurate all-weather precipitation gauge, withspecial emphasis on easy maintenance andextended service intervals. The electronics unitincludes a processor with embeddedalgorithms for calculation of cumulative rainfalland intensity. Versatile interfacing optionsguarantee easy interfacing to various datacollecting systems.

The main features of VRG101 arepresented in table 1.

Gauge type Weighing precipitation gaugeParametersmeasured

- Cumulative precipitation(mm)- Precipitation intensity (mm/h)- Temperature (°C), optional

Sensor element Single point load cellCollecting area 400 cm2 (62 in2)Capacity 650 mm (25 in)Resolution 0.1 mmAccuracy 0.2 mm

- precipitation event > 0.5 mmDimensions Height 950 mm (37 in)

Diameter 400 mm (16 in)Powerconsumption

- Without heating 30 mW- With heating 100 W max

Serial I/O - RS232 and RS485 lines forgauge output andconfiguration.- Polled or automatic message

Data messageparameters

- Gauge status- Cumulative precipitation mm- Precipitation intensity mm/h- Air temperature (option)- Container mass g- Electronics temperature- Supply voltage

Options andaccessories

- Pulse output- Rim heating- Air temperature sensor- Wind shield (types Alter,Double Alter and Tretyakovavailable)- Pedestal- Field check kit

Table 1. VRG101 main features.

2.2 Mechanics

VRG101 utilizes the latest high-accuracy,temperature compensated load cell technology.The single point - type load cell is designed fordirect mounting of the weighing platform.Eliminating levers and flexures, this allowssimple, robust and low cost mechanics.

The load cell is insensitive to eccentricloading unlike some other types of weighinggauges. Thus unsymmetrical distribution ofsnow in the collecting bucket (typical for winterconditions) does not introduce measurementerrors.

The gauge volume and geometry havebeen optimized for good performance in allweather. Its wide 400 cm2 collecting area isadvantageous when measuring light rain andthe large 650 mm net capacity decreases therisk of overflow. The deep container, togetherwith the constriction formed by the inlet funneldecrease the evaporation error and outblowingof collected snow.

Another error source that is eliminated bythe advanced mechanics is theunderestimation caused by water and snowsticking to the inner surfaces of the gauge inletfunnel. In conventional designs this mass is notmeasured and eventually evaporates. InVaisala's design the funnel element rests onthe collector container. All water and snow onit’s surface is therefore included in themeasured mass.

The hinged upper part (rim and collectingfunnel) and detachable enclosure door alloweasy access for maintenance or addingantifreeze solution, as well as easy removal ofthe collector container (see fig. 2).

The electronics unit, including the load cellis field-removable (fig. 3). Replacement of theelectronics is straightforward and quick. Dataloss is kept to a minimum as there is no needto transport the whole gauge to the laboratoryfor calibration. On-field checking of the gaugeaccuracy can be done using a dedicated fieldcheck kit.

Fig. 2. The gauge with enclosure opened.

Fig. 3. Electronics unit and load cell are situatedunder the weighing platform.

2.3 Software and interfacing

The gauge software uses advancedalgorithms to filter out noise, spurious signals(e.g. vibration by wind, mechanical impacts andrubbish or other objects entering the collectingcontainer) and to compensate for evaporation.

The outputs include RS232 and RS485serial lines with polled or automatic messaging.Configuration of the gauge is also donethrough the serial lines. An optional pulseoutput (tipping bucket emulation) withprogrammable tip size is available.

In addition to cumulative rainfall the gaugedata message also includes precipitationintensity, supply voltage, electronics internaltemperature, gauge status with error flags andair temperature (if the optional Pt100 sensor isconnected). Complete raw data (weight of thecontainer) is also available to be used fordiagnostic or research purposes.

2.4 Heating option

Optional rim heating is recommendedwhenever solid precipitation needs to bemeasured. Heating prevents accumulation ofsnow and ice on the rim and the collectingfunnel. To prevent extraneous evaporationerror caused by heating and to minimize powerconsumption, the heating is controlled by thegauge's software. The intelligent controlalgorithm is based on ambient temperature andprecipitation conditions.

2.5 Wind shields

Use of a wind shield is essential for highaccuracy, especially when measuring snow orsleet. There are two types of basic shieldsavailable for the VRG101: Tretyakov and Alter.For climatological stations and otherapplications requiring the highest accuracy, aDouble Alter shield is recommended.

2.6 Other options and accessories

Other options and accessories includepulse output, air temperature sensing kit (aPt100 temperature sensor with a radiationshield), field check weight, gauge pedestal andscrew pole foundation.

Automatic draining option will be availablein spring 2006.

3. TEST RESULTS

The gauge has undergone extensivetesting at Vaisala's own test field, in the FinnishMeteorological Institute's (FMI) observatories inJokioinen and Sodankylä, and by various pilotcustomers. Here the results obtained from FMIfield test facilities and the National AtmosphericDeposition Program (NADP) test field arepresented.

In the FMI Jokioinen Observatory,Southwestern Finland trial, VRG101 wascompared with the FMI's double fenceintercomparison reference (DFIR): a highaccuracy weighing gauge with it's orifice heightset at 3 m and surrounded by an octagonalvertical double fence.

The comparison was started in August2004 and still continues. The results presentedhere are from the period August - November2004 (Turtiainen et. al. 2005). Figure 5 showsthe observed relation between VRG101 andthe FMI reference gauge. Table 1 shows dailytotal values divided into three categories (lightrain, rain and heavy rain).

0,1 1,0 10,0

FMI reference gauge (mm)

0,1

1,0

10,0

Vai

sala

gau

ge (m

m)

R Sq Linear = 0,993

Fig. 5. Daily totals, Vaisala gauge versus FMIreference gauge (DFIR)

Overall, the Vaisala gauge caught slightlymore precipitation compared to the FMI gauge,the difference being highest for light rainevents. The difference is believed to be causedat least partially by the designs of the the inletsections of the gauges. The inlet funnel of thethe Vaisala gauge is a part of the instrumentweighing system, whereas this is not the casefor the reference gauge. Therefore waterdroplets caught on the surface of the inlet ofthe FMI gauge are likely to cause wetting error,

and consequently an underestimationespecially in light rain.

Daily rainfall Nbr ofdays

FMIreferencegauge(mm)

VRG101(mm)

< 1 mm 10 4.5 6.21 - 4.4 mm 15 38.5 43.0> 4.4 mm 12 133.9 138.9TOTAL 37 176.9 188.1

Table 2. Summary of liquid precipitationmeasurements at FMI Jokioinen Observatory,August - November 2004.

Daily rainfall Nbr ofdays

FMIgauge(mm)

VRG101(mm)

< 1 mm 4 2.0 1.21 - 4.4 mm 7 14.1 14.1> 4.4 mm 4 35.4 34.3TOTAL 15 51.5 49.6

Table 3. Summary of snowfall measurementsat FMI Sodankylä Observatory, February - April2005.

Another, identical gauge was installed inthe FMI Sodankylä Observatory in NorthernFinland. Snowfalls are more abundant inSodankylä and extend further into spring thanthey do in Jokioinen. The gauge was equippedwith rim heating and a Tretyakov type windshield. The reference used in Sodankylä was astandard FMI precipitation station: a manualgauge and a Tretyakov type wind shield.

Table 3 shows the summary of the dailyresults. The Vaisala and the reference gaugeswere in good agreement. The difference in thetotal values was only 1.9 mm or 4 %.

VRG101 was also tested by several pilotcustomers. An example of interim results oftests performed at the test field of the NationalAtmospheric Deposition Program (NADP),Champaigne, IL. are presented here. Thesummary of the results is shown in fig. 6. Ingeneral the two gauges were in goodagreement, the difference in the totalcumulative rainfalls being only 2 %, withVRG101 showing slightly more.

4. CONCLUSIONS

The VRG101 belong to a new generationof weighing precipitation gauges. Simple androbust mechanics, optimized gauge geometry,the latest high-accuracy load cell technologycombined with advanced measurement andheating control algorithms ensure highperformance in all weather conditions.

The gauge has been field tested in co-operation with the Finnish MeteorologicalInstitute, as well as with several pilotcustomers. The results demonstrate the goodperformance of VRG101 both in liquid andsolid precipitation.

When interfaced with the Vaisala MAWS-dataloggers and communication modules thegauge can be used both as a basic stand-alonehydrometeorological station, or as acomponent in larger observation networks. Dueto it's versatile output options, the VRG101 canbe interfaced with any data collection systemwith a RS232/485 or a pulse input, and has awide application area in state and national

meteorological, hydrological and climatalogicalnetworks. The precipitation intensity outputalso enables the gauge to be used as areference when weather radar signals are to beconverted to quantitative rainfall amounts.

Acknowledgements. The authors thank Ms AnuPetäjä from the Finnish Meteorological Instituteand Roger Claybrooke from NationalAtmospheric Deposition Program for their kindco-operation in organizing the field tests andpermission to publish the results.

REFERENCES

Turtiainen, H., S. Räisänen and A. Petäjä2005: Next generation all weather precipitationgauge. WMO Technical Conference onMeteorological and Environmental Instrumentsand Methods of Observation (TECO-2005).Bucharest, Romania, 4-7 May 2005.

0.94

0.46

0.00

4

0.03

0.39

1.69

0.25 0

.45

0.99

0.47

0.00

0.03

0.40

1.72

0.26 0

.46

0.00

0.50

1.00

1.50

2.00

8/19

/200

5

8/20

/200

5

8/22

/200

5

8/26

/200

5

9/14

/200

5

9/16

/200

5

9/19

/200

5

9/20

/200

5

Rai

nfa

ll (in

)

NWS stick gage VRG101

Fig. 6. Comparison of daily rainfall measured by VRG101 and NWS stick gage at the NADP test field,Champaigne, IL, from 19 August to 20 September 2005.