An automatic device for measurement of the amount (weight) of deposited precipitation developed at the Institute of Atmospheric Physics, Academy of Sciences of the Czech Republic, is described. Examples of measurements of various types of deposited precipitation are presented. The paper also discusses the response of the measuring instrument to falling precipitation and the influence of wind on the measurements. The results of first measurements proved that the instrument is suitable for automatic and continuous monitoring of deposited precipitation.
1. Monitoring of deposited precipitation
Deposited precipitation, such as dew, frozen dew, white frost, and fog, represent an addition of water to the ecosystem that is very difficult to quantify. Together with water from deposited precipitation, polluted matter that is contained in the atmosphere near the surface also enters the ecosystem. Many scientists are concerned with the pollution of deposited precipitation (e.g., Acker et al. 2002; Aikawa et al. 2007; Fišák et al. 2001). There is a lack of data regarding the amount of deposited precipitation required for determination of the deposition. Various estimates based primarily on liquid water content have been used (e.g., Kunkel 1984; Fiser 1996; Fisak et al. 2006). Presently, measuring deposited precipitation is never, or very rarely, performed and includes only a few types of deposited precipitation.
In the past, attempts to measure deposited precipitation have resulted in the production of a few prototypes of measuring devices. Various absorption boards (most frequently made from plaster) were used. These boards were dried and weighed before the experiment. After the experiment, they were weighed again, and the difference between the two weights represented the weight of the dew (deposited precipitation). This method of measuring the deposited precipitation weight was very laborious and often unreliable. The results of the measurements were fully dependent on the precision and reliability of the experimenter.
Attempts to develop registration devices have also been performed, such as the Kössler dew meter. These devices operated on the principle of mechanical scales and used mechanical registration devices with a clock and a registration pen. The retaining area slid out from the common register device and represented one beam of a balance. The registration pen was protected against the wind using a special oil shock absorber. These devices were quite expensive and unreliable when the temperature dropped below 0°C. The mechanical transmissions and oil shock absorbers sometimes hardened. Detailed descriptions of the various methods for dew measurement are reviewed by Uhlíř (1948).
The only widespread device for determining dew amount was the Duvdevani dew meter (Duvdevani 1947), a varnished wooden prism of the required dimensions that was located in various heights above a clipped lawn (from ~5 cm to 1.5 m). The quantity of the dew is determined by the Duvdevani dew measuring scale (Stružka 1956; Middleton and Spilhaus 1953). This scale is represented as images of the dew, and the corresponding amount of dew measured in millimeters of precipitation is indicated next to every picture. However, using this device allows the experimenter to subjectively estimate only the quantity of dew for which this scale was developed. Importantly, this scale was not developed for different types of deposited precipitation.
Thus, the measurements or estimates of dew amount are dependent on the human factor. These measurements must be performed as close to sunrise as possible, as that is when the maximum amount of deposited precipitation is observed. After sunrise (depending on the season), the deposited precipitation begins to vaporize. This means that, when the measurement is delayed, the data can be skewed or, in an extreme case, the dew is not detected at all. Krhounek (1956) and Krečmer (1958) pointed out some issues related to the use of this method.
In spite of these issues, the Duvdevani method spread in the 1950s and 1960s and was used at the meteorological stations of the Hydrometeorological Institute in Czechoslovakia. In the 1970s, measurement of deposited precipitation was gradually terminated. At present, meteorological stations only provide information about the presence of deposited precipitation and its intensity in the range between 0 (weak phenomenon) to 3 (very strong phenomenon)–(Fišák 1994). The quantity of deposited precipitation is recorded only when precipitation is strong enough (quantities of 0.1 mm or more) to be observed by a rain gauge. The only exception is solid deposited precipitation. In 2000, a device for measuring rime quantity was developed at the Institute of Atmospheric Physics (IAP), Academy of Sciences of the Czech Republic, and is described by Fišák et al. (2001). This device is used for the present investigation but not for routine operation.
Dew, its chemical characteristics, and the condensate dampness amount were studied by Beysens et al. (2001), Muselli et al. (2002), and Beysens et al. (2003). These authors, among others, mention devices for sampling and registration of deposited precipitation. The device described in this paper targets only the registration of the weight of deposited precipitation without the demand on deposited precipitation sampling. This allowed for both easier solving of and reduction of the sampling area.
2. Measuring instrument
A new device for the measurement of the amount of the deposited precipitation (dew) was developed at the IAP. The device measures the actual weight of the precipitation deposited on the collecting board. A diagram of this device can be found in Fig. 1. It is composed of two main parts: the case containing the electronics and the collecting board.
The case containing the electronics (1) is attached to a pedestal (7) that provides sufficient stability for the device. The connector (2) attached to the case is used to supply power to the electronics and for communication with a computer.
A thermally compensated tensometric bridge and electronic system are located in the electronics case. The bridge is connected by the holder (3) with low thermal conductivity to the collecting board (5). The opening for the holder in the case is protected from insects and dirt by a gasket cover (6). To reduce the impact of heat from the electric circuit to the collecting board, a thermal cover (4) is placed above the case containing the electronics.
The analog part of the electronics consists of an accurate voltage reference that feeds the tensometric bridge and analog power supply for an accurate 24 bit sigma–delta analog-to-digital (AD) converter. The output of the tensometric bridge is tied to the input of the AD converter. Digitized measured values are read from the AD converter by the microprocessor. The measured data are given in 0.5-g resolution. The stability of the output value was tested in the temperature chamber; it was better than ~2 g over a temperature range from −20° to 60°C. The collecting board has an area of 0.2 m2. The measured weight of precipitation is usually recalculated to the height of precipitation per 1 m2. Thus, the resolution is ~0.0025 mm of water per square meter. The collecting board is made from Plexiglas (reinforced by metal arms from the bottom) and is set to a horizontal position by adjusting the terrain, as the tested device is not equipped with setting screws to minimize the production costs. The bottom side of the collecting board is covered by an aluminum reflecting layer to prevent the ground below the collecting board from warming by solar radiation and, hence, to prevent heating of the collecting board by the secondary longwave thermal radiation from the ground surface below. At the same time, this reflecting layer constitutes a part of the thermal shield that minimizes the impact of the heat loss from the electronics of the device on the measurement of dew weight.
The device was tested at the Station for Transport Processes and Soil Moisture Dynamics Monitoring of the Department of Water Resources, Faculty of Agrobiology, Food and Natural Resources, Czech University of Life Sciences in Prague (DWR FAFNR CULS). The station is located inside the area of the university on the north-northwestern edge of Prague in the town district Suchdol. There is only agricultural land surrounding the station where the device is installed, and low, single-story, agricultural buildings are located farther than 100 m from the device. The location of the DWR FAFNR CULS station and the dew meter placement are shown in Fig. 2.
The described device is currently being tested at the DWR FAFNR CULS and has been in operation since 1 March 2011. Further manipulation of the device is expected because its position is temporary: there is a wind influence at its temporary location. The definitive installation of the device will involve placing the device into a hole with protection against precipitation leakage and mounting the collecting board at a height of 5 to 10 cm above the terrain. This placement should also considerably reduce the impact of wind.
The new device continuously records the weight of the deposited precipitation every 10 min, analogous to other meteorological measurements performed at the station (i.e., air temperature, visibility, wind speed, and wind direction, all at a height of 2 m).
Figure 4 shows a typical record of deposited precipitation with the progression of relative air humidity (RH) at 2 m. This record is similar to the record in Beysens et al. (2003). In this paper, the focus was mainly on the presentation of precipitation behavior and some of meteorological phenomena on the device record. The dependence of deposited precipitation amount and meteorological conditions is viewed only marginally. This topic will be studied in another paper, where the focus will be mainly on the comparison of deposited precipitation amounts measured by the observer with the help of the Duvdevani dew meter with amounts measured by our new device.
The present study focuses on the measurement of the amount of deposited precipitation by the new device. The paper includes results from measurements obtained between March and August 2011.
3. Saving and presentation of measurements
As previously mentioned, the measured values of deposited precipitation are saved on the device every 10 min. The device can operate both on- and offline. For offline use, it is possible to download the data from the inner memory of the device (capacity for ~50 days) to a notebook. For online use, the device can be connected to a computer and data can be downloaded via a computer network at a distant terminal. The data are saved to a text file that can be easily read and transformed to a data or Excel file. Afterward, the data can be displayed as a graph. Figure 3 shows a record of values taken during the first month of measurement using this device (March 2011). Measured values can be negative when the collecting board is disturbed by the wind. The negative values are usually suppressed, and only positive values are displayed. Figure 4 shows a slow linear increase in the amount of white frost that lasted until 0840 Central European Time (CET). After that, a strong decrease due to evaporation followed. Sunrise occurred at 0625 CET 5 March 2011. Thus, the amount of white frost was still increasing 2 h, 15 min after sunrise. After 3 h, all of the white frost had evaporated.
Apart from precipitation deposited from fog, which can arise at any time during the day, the increase in amount of deposited precipitation is traditionally highest at night, so the total amount peaks around sunrise. Deposited precipitation is attributed to the day when its maximum was reached, and this maximum is taken as the amount of precipitation recorded for the specific event.
4. Recording of falling precipitation
Other phenomena were recorded along with deposited precipitation, namely, various types of falling precipitation and wind gusts. In the case of falling precipitation, the phase of the precipitation is important. If the amount of liquid falling precipitation exceeds ~0.2 mm per 10 min, water starts running off the collecting board, and the total amount cannot be recorded. On the other hand, the new device is capable of recording a small amount of falling liquid precipitation not measurable by a rain gauge. It is also possible to measure the amount (weight) of falling solid precipitation, as it will stay on the collecting board.
Figure 5 shows the occurrence of falling precipitation (in this instance, rain) and clearly shows the different characteristics of recording this type of precipitation compared with those of deposited precipitation. Therefore, it is clear that it is possible to separate the measurements of falling and deposited precipitation. From 1010 to 1340 CET, 3.9 mm of precipitation fell. Because the precipitation amounts exceeded 0.2 mm in a time span of 10 min, the water that fell on the collecting board during this time was not measured, and the record shows smaller values of precipitation than the amount that actually fell. The collecting board gradually dried after the precipitation had ended. Precipitation amounts per 10 min measured by a rain gauge are shown as the dashed line labeled SRA in Fig. 5. The impact of wind flow on the collecting board can also be seen in the record.
The deposited precipitation amounts from 0610 CET 16 March 2011 to 1230 CET 18 March 2011 are reported in Fig. 6. In the time from 1050 CET 17 March 2011 to 1230 CET 18 March 2011, a total amount of 16.3 mm of precipitation fell. The precipitation was mixed (including rain and snow) but gradually changed to snow, which then slowly melted. On 16 March 2011, a very weak rain occurred in the early morning, which was not detected by the rain gauge. Even here, one can see that the record significantly differs from the record of deposited precipitation.
Figure 7 shows how very weak falling precipitation can be detected by this device. The quantity of this precipitation was less than 0.05 mm. In the daily records of the meteorological station, such precipitation is reported as immeasurable.
5. Recording of wind flow
Figure 8 shows the vibration of the collecting board due to wind. In previous figures the negative values were intentionally omitted, but are retained here. The figure shows that the collecting board is both lifted and weighted down by the wind. The impact of wind flow was not eliminated for the following reasons.
The increase of deposited precipitation is usually highest when wind flow is negligible.
Only a small set of data exist at this time.
The amount of deposited precipitation is defined as the highest measured value.
The impacts of wind flow are so apparent in the record that it is not difficult to identify them.
Precipitation deposited from fog is recorded.
At the time of testing the new device, three episodes of fog occurred. It is necessary to note that these were all cases of radiation fog, when the fog originates as a consequence of radiation cooling of the air. To investigate fog precipitation deposition, the longest fog episode, which lasted from 0310 to 0830 CET, was chosen. Figure 9 shows the measured deposited precipitation (solid line) and visibility (dashed line). When the fog appeared (in this case, the visibility fell to less than ~200 m), the deposited precipitation started to increase more rapidly. After reaching its peak, the value did not decrease rapidly like in the cases of dew or white frost. As previously stated, the values were obtained during a period of radiation fog, so the record was similar, to a certain extent, to the records of dew or white frost. The same process occurred in the atmosphere on a different scale and not only in the layer near the surface. Many kinds of fog exist, which are defined by their origin (Sobíšek 1993). Different types of fog include radiation fog, which originates from radiation cooling of the air; advection fog, which originates from the cooling of relatively warm and humid air by its advection over a colder surface; advective radiation fog, which consists of a combination of radiation and advection fog; frontal fog, which is connected to atmospheric fronts and advective changes in temperature and air saturation by frontal precipitation; urban fog, which originates around big industrial cities as an effect of air pollution; upslope fog, which is a consequence of the adiabatic cooling of air moving up a mountain slope, originating on windward sides of mountain ranges; and valley fog, which occurs in terrain depressions as a consequence of colder air flowing down the valley slopes. While there are other kinds of fog than those described above, it is likely that deposited precipitation from different kinds of fog would be recorded differently.
6. Deposited precipitation evaluation
To evaluate recordings of the amount of deposited precipitation weight, other meteorological data such as falling precipitation, wind speed, visibility, air temperature, or humidity are needed. The typical record for deposited precipitation should be chosen. When specific records are not clear, other meteorological data can be used. The temporal change of air temperature helps to determine the phase (type) of deposited precipitation. In this case, the surface temperature approximately 5 cm above the terrain is most suitable. If such data is unavailable, it is possible to use the temperature at 2 m above ground.
The most convenient process of evaluating the amount of deposited precipitation is as follows:
The development of the amount of deposited precipitation is depicted.
That part of the record that shows characteristic development of deposited precipitation is chosen.
Relevant maximal values of the amount of deposited precipitation in the dataset are obtained.
Records that exhibit atypical characteristics are analyzed. It is then determined whether a given phenomenon deposited precipitation or not using additional information like air humidity, occurrence of falling precipitation (from rain gauge), or, alternatively, information from a precipitation detector, if necessary.
When information about the phase of deposited precipitation is also required, the temperature at the time of deposited precipitation occurrence must be considered.
Examples of the evaluation of deposited precipitation collected in March 2011 were presented in the paper. Figure 3 shows 16 cases of deposited precipitation recorded in March 2011. In one case, the measurement was impacted by increased wind flow, and in another case, the deposited precipitation increased when the collecting board did not have sufficient time to dry following weak falling precipitation. In March 2011, the total deposited precipitation was 1.46 mm, which corresponds with 8.9% of falling precipitation.
Owing to the relatively short time of operation of the device, we could not record the deposited precipitation of all possible phenomena. For example, we have not obtained measurements of deposited precipitation from several different types of fog. Various forms of solid falling precipitation, like snow or hail, were also not measured. Therefore, it is not clear how the device will react to these phenomena. We suppose that in some cases the collecting board should be cleaned at a specified time (0700 CET), as in the determination of new snow cover.
The main target of this study was to find if the new device was useful for deposited precipitation measurement. Even after a short time of observation, it is evident that the new device is more sensitive and records more days with deposited precipitation than observers at meteorological stations. It was shown that the new device is suitable for continuous measurement of deposited precipitation. The preliminary results indicate that deposited precipitation at the Prague–Suchdol station can reach up to 10% of the amount of falling precipitation. It is anticipated that this value will change based on the altitude of the station where measurements are obtained. For now, with the very low number of available measurements it is not possible to make definite conclusions. New measurements are currently in progress, and the results will be processed in the next study. In the future, we will also focus on the contribution of deposited precipitation to the amount of falling precipitation.
The results described in this paper were obtained in the frame for AS CR with the support of the GACR Project 205/09/1918 and the investigation projects V0Z AV0Z30420517 and MSM6046070901.