Лёд и Снег • 2012 • № 4 (120)
УДК 551.321.4
Snow temperature measurements at Vostok station from an autonomous recording system (TAUTO): preliminary results from the first year operation
© 2012 г. E. Lefebvre1, L. Arnaud1, Л.А. Ekaykin2, V.Ya. Lipenkov2, G. Picard1, J.-R. Petit1
1Laboratoire de Glaciologie et Géophysique de l'Environnement, CNRS/UJF Saint Martin d'Hères, France;
2Arctic and Antarctic Research Institute, St. Petersburg elec@lgge.obs.ujf-grenoble.fr
Статья принята к печати 17 октября 2012 г.
Antarctica, autonomous recording system, snow temperature, snow thermal properties, temperature monitoring, winter warming. Автономная регистрирующая система, Антарктида, зимнее потепление, мониторинг температуры, температура снега, термические свойства снега.
Temperature gradients in the upper layers of the snow pack are of importance for studying the emissivity properties of the snow surface with respect to microwaves used in remote sensing as well as for the heat and mass transfer in snow thickness. Gradients drive the initial snow microstructure metamorphisms that probably influence the firn properties in regard to air molecules fractionation and the air bubble enclosure process at close-off depths. As a contribution to investigation of these problems and following J.-M. Barnola initiative, we developed an autonomous recording system to monitor the temperature of the upper layers of the snow pack. The instrument was built to be autonomous and to be continuously operating within environmental conditions of the Antarctic plateau and the polar night. The apparatus which monitors temperature from the first 10 m of snow by 15 sensors of a «temperature grape» was set at Vostok station during 55th Russian Antarctic Expedition within the frame of the French Russian collaboration (GDRI Vostok). From the available hourly measurements over the first year, we present preliminary results on the thermal diffusive properties of the snow pack as well as some character of the temperature variations on the Antarctic plateau.
Introduction
Polar ice core provide a wealth of information on climate and on its dynamics over period of time encompassing by now the last 0.8 million years [6]. The most salient result from the ice core studies is the close correlation between temperature and greenhouse gases (CO2, methane) firstly observed from the Vostok ice core over one then four climate cycles [2, 17]. This correlation is now extended over the last 0.8 million years [13, 21], and the information that present-day greenhouse gas levels are unprecedented high had a large impact on several IPCC assessments on Global Climate Change [10].
While climate proxies (e.g. water isotopes, chemical elements, dust...) are imprinting the upper snow layers at a given time, gases are entrapped in air bubbles 50 to 120 m below surface at the close off depth and within layers 1000 to 6000 year older. The depth and age difference estimations are very challenging as they depend on densification processes which are variable with site temperature and accumulation [9, 15, 19] but also from various factors. Among observations, the ice core records of O2/N2 ratio and the air volume contains signal variability at orbital frequencies [3, 18] suggesting the firn properties in depth may keep memory from solar radiation in initial stage of metamorphosis of upper snow layers [12]. Surface snow layers are very porous and are subject to metamorphosis due to combined effects of solar radiation, weather conditions and
strong daily temperature gradients. Transfers of heat, water vapor and gases are intense and they contribute to firn ageing and post-depositional effects with fractionation of water isotopes [4, 5, 11], the photolysis of chemical elements as nitrates within the very upper layers [7], and the gravitational fractionation of gases in depth [20].
Temperature and properties of the upper snow layers are also of importance with respect to the remote sensing observations of Polar Regions, for interpreting the brightness temperature time-series which are acquired routinely by passive microwave [16]. Of interest is the 19 GHz brightness temperature which is sensitive to snow temperature at depth, up to a few meters [22] but also from the microwave penetration depth which is unknown. In this respect, in-situ temperature measurements for the upper layers at a site represent a «field truth» and will help for modeling the emissive properties of the snow pack.
As a contribution to these problems, and following a project promoted by J.-M. Barnola (deceased 2009); we developed an autonomous system for continuous recording the temperature variations of the upper layers of the snow pack to be deployed at some selected sites in Antarctica. One instrument was deployed at Vostok station in the frame of the Russian French collaboration (GDRI Vostok). Here we present the data from the first year of operation along with preliminary values for thermal diffusivity and some character of the temperature variations during winter.
Fig. 1. TAUTO recording system and its setting in the field. Top - insulated box containing the central electronic unit, and details of the electronic cards with low energy consuming microprocessor; bottom -setting the instrument in a pit at Vostok station during summer 2010 (RAE 55) and data downloading operation during summer season (RAE 56) Рис. 1. Автономная регистрирующая система TAUTO и её установка в полевых условиях. Вверху — термоизолированный бокс, вмещающий основной электронный блок и элементы электронной карты с энергосберегающим микропроцессором; внизу - установка системы в шурфе на станции Восток летом 2010 г. (55-я РАЭ) и скачивание накопленных данных в летний полевой сезон 56-й РАЭ
Methods and results
The TAUTO recording system. Autonomous and automatic digital recording system was built at Laboratoire de Glaciologie et Géophysique de l'Environnement (LGGE) and customized for operating within the low temperature environment of the Antarctic plateau. The «Temperature AUTOmatic (TAUTO)» recording system uses a low power consummation microcontroller which scrutinizes voltage drop from up to 64 platinum resistances (PT100). It has capability for data storage for years on flash disk and for transmitting the data to ARGOS communication system satellite. The electronic hardware is composed from three main cards: the central unit, the memory and the transmitter (Fig. 1). All is fitted for operating by -40 °C, and placed in a sealed insulated box buried in snow. Miniaturized heating systems allow pre-heat-ing of the most temperature sensitive electronic chips during a programmed short time before the cycle of measurements. The power supply is given by two 10 watts solar panels set at surface and six 60 Ah buffer batteries buried in snow to prevent the very low temperature during winter. The system is fitted to operate in full autonomy of energy during 4 months and to overcome the polar night.
Attention was paid to configure the temperature probes. The manufactured Platinum Resistance Temperature (PRT) detectors have been chosen for both the relative small size and the accuracy (IEC751 (1983), 1/10 DIN) of ±(0.03 + 0.0005-|71) °C. This DIN standard interchangeability for Pt resistance temperature detector is only for the uncounted sensor, and is higher for mounted probes (Pt sensor with metallic sheath, 4 lead wires, and a connector). So, we measured the effective relative accuracy of the mounted PRT probes: ±0.15 °C on the range of our measurements (-20 °C to -90 °C).
Each of the 15 sensors was connected to a small diameter (1.5 mm diameter) 4 wires Teflon® line (white color) of various lengths (up to 15 m) and assembled together to make a so-called «temperature grape». The grape is connected to the main unit.
Improvements of sensor calibration were conducted at LGGE in order to reduce the interchangeability of the PT 100. For this relative inter-calibration, the 15-sensors from the grape were measured all together with one reference temperature sensor (accuracy of ±0.05 °C with respect to absolute temperature) at 3 different temperatures (—30, -45 and
-58 °C). To a given stable condition, the 15-sensors mean resistance value was calculated. A correction factor was applied to each resistor in order to provide a resistance value closest to the mean stack value. A tabulated version of the Callendar-Van Dusen equation is used to calculate temperatures from resistances. This calibration stage between all sensors resulted after corrections with a relative accuracy of ±0.05 °C and an absolute precision better than ±0.1 °C (comparison with our calibrated reference probe).
Each hour the system wake-up automatically from a real time clock (RTC), and after 2 mn of warm-up, a new cycle of measurements starts. Time is tuned from a GPS receiver. A low excitation current (1 mA) is applied to PT100 through (2 wires) for minimizing self-heating of sensor while the voltage drop across the sensor is measured on the 2 other wires by a high impedance amplifier. The 4 wires resistance readings are multiplexed one by one during 1 sec. The analogic signal is converted to digital values then to resistance value (and temperature). The data are stored in an 8 Mo flash memory and transmitted to satellite (ARGOS system). ARGOS data are weekly downloaded to LGGE and to AARI from the reception center in Toulouse (France). The power consumption is 60 mW as background, then 100 mW during 2 mn every hour to scan sensors for temperature measurements, then 30 mW for 2 mn twice a day for a GPS receiver used for the time synchronization, and finally 250 mW during 2 sec every 200 sec of the emission to ARGOS. The redundancy of emission is 2 or 3 times and reduces download errors.
The whole system has been set in 2010 (55th Russian Antarctic expedition) within the clean sector of Vostok station. A 3 m deep pit was dug for burring the electronic unit and batteri
Для дальнейшего прочтения статьи необходимо приобрести полный текст. Статьи высылаются в формате PDF на указанную при оплате почту. Время доставки составляет менее 10 минут. Стоимость одной статьи — 150 рублей.