New concept of permafrost degradation monitoring based on photonics technologies: Case study from Calypsostranda (Bellsund, Svalbard)

Ground temperature measurements are crucial for a better understanding of changes in the natural environment, especially in the Arctic. Previous measurement systems provided accurate measurements; however, their most significant disadvantage was the relatively low spatial resolution, including in the vertical profile. The aim of this work was to develop and initially validate a new, original temperature measurement system based on the photonic sensing technique of optical frequency‐domain reflectometry (OFDR). The system consists of a fibre‐optic sensor, an interrogator, and an automatic data acquisition system. Such fibre‐optic sensors allow a significant increase in spatial resolution. Data on precise temperature distribution in the ground profile will allow for a detailed determination of the changes in the thickness of the permafrost active layer (PAL) and, as a consequence, a better description of the current state of the permafrost and the layers above it in relation to their progressive degradation. In the longer term, it will make a better prediction of the pace of possible changes in the polar environment and will open up previously unavailable opportunities in the field of climate change monitoring and forecasting.


| INTRODUCTION
The High Arctic is experiencing rapid warming, leading to declining glaciers, degrading permafrost, and ultimately rising sea levels and coastal erosion (Etzelmüller et al., 2020;IPCC, 2014;Rantanen et al., 2022;Romanovsky et al., 2017;Stern et al., 2012).The main problem is especially the increasing dynamics and range of changes in the permafrost active layer (PAL) that may affect the stability of slopes, ecosystems, and infrastructure (Harris et al., 2009;Hjort et al., 2022;Jask olski et al., 2018).They can also disrupt the stability of organic matter, causing the emission of greenhouse gases stored in it (Canadell et al., 2021;Miner et al., 2022).
The amount of thermal energy supplied to the ground determines the depth of soil thawing (permafrost) and the thickness of the PAL.
The thawing of permafrost in a warming climate is governed by a complex interplay of different processes, of which conductive heat transfer seems to be the most important.The compactness and stability of the frozen ground are closely related to its temperature, making monitoring this quantity essential for assessing the stability of the local landscape and quantifying the impact of climate change on cold regions (Christiansen et al., 2019;Dobi nski & Kasprzak, 2022;Haeberli et al., 2017;Jones et al., 2019;Walvoord & Kurylyk, 2016).
The dynamics of the PAL depend mainly on climate conditions (Christiansen et al., 2003;Repelewska-Pekalowa et al., 2013).Its variability is also determined, to a certain extent, by local factors, such as terrain configuration and aspect, vegetation cover, as well as the kind and degree of ground-water mobility (Brown et al., 2000;Christiansen et al., 2003;Repelewska-Pękalowa & Gluza, 1988).
Data on the precise temperature distribution in Arctic regions, both of the subsoil and the ice sheet, are crucial for understanding the causes of climate and environmental change, providing a basis for a diagnosis of the current state, and for understanding the phenomena, which are currently occurring, as well as predicting their consequences.Research to date shows that Svalbard, in particular, faces major challenges.Climate change is happening here even faster, and the changes are more visible than in other Arctic areas (Ding et al., 2018;Liljedahl et al., 2016).Currently, large-scale monitoring is being carried out in Svalbard, the main purpose of which is long-term documentation of the distribution, condition, and changes to permafrost (Biskaborn et al., 2019;Christiansen et al., 2019;Smith et al., 2022).In many parts of the archipelago, this research has a relatively long history.This also applies to the Calypsostranda region, where PAL dynamics have been monitored since the mid-1980s (Repelewska-Pekalowa et al., 2013;Repelewska-Pękalowa & Pękala, 2004).Despite the intensification of research on the degradation of the Spitsbergen permafrost, the quality and quantity of data is still not satisfactory, and their availability is limited.That is why it is crucial for a detailed recognition of the process and prediction of its effects in the near future to design effective systems for obtaining quick access to such data.The key technological solution for this issue is the use of a new generation sensor to study the thermal properties of the ground in permafrost conditions, giving the possibility of precise, high-resolution measurements in a relatively short time interval.
So far, ground temperature measurements have been based on thermistors or thermocouple sensors (Christiansen et al., 2010;Noetzli et al., 2021).Such measurement systems feature relatively low spatial resolution.This low resolution means that it is impossible to precisely monitor the processes taking place in the PAL, especially on the border of the permanently frozen and thawed zones.
In recent years, fibre-optic sensors, which offer the possibility of performing truly distributed measurements, have become quite popular for geophysical and hydrological surveys (Blanc et al., 2022;Fenta et al., 2021;Gorshkov et al., 2022;Schenato, 2017;Selker et al., 2006;Shanafield et al., 2018;Tyler et al., 2009).Such photonic sensors have also been considered for applications in permafrost environments.For example: (1) Raman-scattering-based distributed temperature sensing (DTS) installed along a transect on the surface of Helen Creek rock glacier has been used to measure, with a spatial resolution of 1 m, the bottom temperature of snowpack in parallel to discrete electronic thermometers (Harrington & Hayashi, 2019); (2) DTS has been installed in permafrost to monitor risk to integrity of a road resulting from permafrost thawing (Roger et al., 2015); (3) thermal resolution and accuracy of measurements performed with DTS has been evaluated in a 100-metre-deep borehole in French Alps (Schoeneich et al., 2015); (4) DTS was also used for measurement of thermal conductivity in deep (few hundred metres) boreholes in Arctic Canada (Henninges et al., 2005;Stotler et al., 2011); however, according to our best knowledge, fibre-optic distributed sensors based on Rayleigh scattering and optical frequency-domain interrogation technique, which offers much better spatial resolution on the order of 1 cm compared to 25 cm of the state-of-the-art Raman-scattering-based DTS, have never been used for temperature monitoring in permafrost.
Recognizing this metrological gap, the authors of this work aimed to develop and initially validate the new, original temperature measurement system based on Rayleigh scattering in fibre-optic distributed sensors and optical frequency-domain interrogation technique.
We hypothesized that such a system make an important contribution to the creation of a new generation of fibre-optic sensors that would monitor ground temperature changes with high resolution at any time and depth interval.

| Previous methods of ground temperature measurement
Mercury ground thermometers have been the basic metrological instrument since the inception of ground thermals measurements and determination of the extent of permafrost.They are, however, extremely sensitive to environmental conditions and can only be mounted at specific points to a certain depth.Obtaining measurement data required reading temperature values from thermometers at specific times.Technological progress in the field of microprocessors has resulted in the creation of new electronic measuring devices owing to which digital recording of measurement data has become possible.
In the last few decades, the most common method of measuring temperature in a permafrost profile has typically involved a chain of thermistors (coated or encapsulated in plastic tubing) or thermocouples placed in a hole drilled in the permafrost and connected to a datalogger (Christiansen et al., 2019;Isaksen et al., 2022).Digital sensors, where a signal is digitized on the spot, are the alternative to this method as opposed to analogue thermistors (Christiansen et al., 2010;Noetzli et al., 2021).For temperature measurement in dry boreholes, an infrared sensor can also be used (Junker et al., 2008).Preparing a series of sensors to cover the measurement profile as fully as possible requires installing multiple sensors, which makes such a system relatively expensive.This creates the problem of obtaining a satisfactory measurement resolution in the profile.Another drawback is that a sensor is inherently quite a complex electronic device, making it vulnerable to damage in extreme permafrost conditions.Each subsequent sensor also increases the requirements for a faster datalogger and power supply in general.The digital temperature sensors strings in which the signal is digitized at the measurement point reduce inaccuracies caused by interferences and signal transmission on long distances.They are also less prone to moisture and are, however, usually installed on a single cable, so if one sensor fails, the entire array stops working (Noetzli et al., 2021).

| The main assumptions of the distributed sensor concept
Photonic technologies give unprecedented possibilities for temperature measurement on the ground in the form of truly distributed sensing, which is not offered by any other branch of measurement tools, such as analogue thermometers or electronic sensors.With optical fibres used as sensors, each along the point of a single optical fibre might serve as a sensing point that can be separately addressed and located, resulting in a continuum of thousands of data points (Blanc et al., 2022;Gorshkov et al., 2022;Palmieri et al., 2022).
Distributed fibre-optic sensing techniques are based on one (or more, in case of hybrid implementations) of three optical scattering phenomena-namely Rayleigh, Raman, and Brillouin scattering.There are many interrogation methods that enable the readout of the distributed data on the physical field of interest using reflectometric interrogators.These interrogators conveniently connect to just a single end of a sensing optical fibre and make use of the backscattered portion of light only (Blanc et al., 2022).
Most commercially available interrogation devices based on Rayleigh scattering (such as OTDR-optical time-domain reflectometer and φ-OTDR-phase-sensitive OTDR, also known as DASdistributed acoustic sensor and DVS-distributed vibration sensor), as well as on Raman DTS and Brillouin (BOTDR/BOTDA-Brillouin optical time-domain reflectometer/analyser also known as DTSSdistributed temperature and strain sensor) have spatial resolution of an order of meters (Fernández-Ruiz et al., 2022;Li & Zhang, 2022;Lu et al., 2019;Palmieri et al., 2022;Schenato, 2017;Silva et al., 2022); however, a Rayleigh-scattering-based optical frequency-domain reflectometer (OFDR), based on the same measurement technique as certain implementation of light detection and ranging (LiDAR) (Ding et al., 2018), namely frequency-modulated continuous-wave (FMCW), has commercially available implementation that allows for a spatial resolution of 1 cm for sensing range of 70 m (Liang et al., 2021).
Although in recent years, DAS, as well as Raman-and Brillouinscattering-based distributed sensors, have been quite a popular choice for geophysical and hydrological studies (Blanc et al., 2022;Fenta et al., 2021;Gorshkov et al., 2022;Schenato, 2017;Selker et al., 2006;Shanafield et al., 2018;Tyler et al., 2009), Rayleighscattering-based OFDR seems to be best-fit for temperature profiling of grounds (Schenato, 2017).It can provide desired dense and abundant information on temperature distribution that can reveal new insights into processes and provide sufficient input data for simulations (Palmieri et al., 2022;Selker et al., 2006).Only some interrogation devices based on Brillouin scattering can rival the number of sensing points and spatial resolution provided by this method (Shanafield et al., 2018); however, those have not yet been proven in commercial implementations.Additionally, due to the low efficiency of Brillouin scattering (both Raman and Brillouin scatterings are much weaker than Rayleigh scattering), these devices are complex and have a disadvantageous additional requirement of connecting to both ends of the sensing fibre (Coscetta et al., 2020;Lu et al., 2019;Palmieri et al., 2022).

| STUDY AREA, MATERIAL, AND METHODS
The study area is situated in the southwestern part of Spitsbergen (Wedel Jarlsberg Land), which belongs to the Svalbard Archipelago (Figure 1a,b).The thickness of the permafrost in this region ranges from 200 to 450 m (Christiansen et al., 2003;Kristensen, 1988;Landvik et al., 1988;Liestøl, 1976).
PAL dynamics measurements in this area have been carried out since the mid-1980s as part of subsequent Polar Expeditions of Maria Curie-Skłodowska University.The tasks performed by successive research teams in this area included studying the dynamics and depth of permafrost melting during the summer and identifying local factors that influence this process.Three methods were employed for this purpose: (1) steel rod probing, (2) Daniilov freezing probing, and (3) determination of the depth of the 0 C isotherm in the ground using either a soil thermometer or a temperature gradient probe.
The maximum extent of summer thaw exhibited significant variability, ranging from 40 cm to over 225 cm (Repelewska-Pekalowa et al., 2013).Based on published data, it is evident that between 1986 and 2002, despite year-to-year variability (ranging between 0.9 and 1.45 m), no clear trend in changes in the PAL thickness can be identified (Repelewska-Pekalowa et al., 2013).From 2002 to 2004, however, a strong and distinct tendency was observed for the PAL thickness to increase during the period of maximum summer thaw

| Geological, morphometrical, and geophysical survey
For a detailed recognition of the lithological context of the study area, core geological drillings and electrical resistivity tomography (ERT) profiling were performed (Figure 3a-e).Along the planned measurement transect, five boreholes were drilled at intervals of 20-50 m using the Eijkelkamp percussion drilling set for heterogeneous grounds with an RD32 (core sampler diameter of 5.0 cm) (Figures 1d   and 3c).The drilling points were geodetically located using a GPS device.
The photogrammetric and morphometric study of the research area was prepared on the basis of vertical RGB photos taken by a Phantom 4 DJI drone (unmanned aerial vehicle, UAV), equipped with the FC631OS camera.The drone was flown at a height of 30 m above the ground.It took place in the GPS navigation mode, so it was necessary to prepare control points in the field.Their position was determined using precise global navigation satellite system GNSS receiver, Topcon Hiper HR, relative to the reference station (CALY point).The data obtained in this process were then subjected to photogrammetric processing in the Agisorf Metashape Professional environment, and all data were imported to the WGS84/UTM zone33 system.The result was a digital elevation model (DEM) with a pixel resolution of 2.07 cm and an orthophoto map with a pixel resolution of 1.04 cm.The visualization of the result was created using the Global Mapper Pro v23 program (see Figure 3a,b).
Noninvasive ground profiling using ERT was carried out along the transect-covering the installation zone of the fibre-optic sensor (see Figure 3d,e).Two ERT models were made along the transect: (1) with an electrode spacing of 4 m and a depth of approximately 25 m, and (2) with an electrode spacing of 1.25 m and a depth of approximately 15 m.Specialized Ares II equipment from GF Instruments, s.r.o, Geophysical Equipment and Services was used for this purpose, and The Wenner-Schlumberger configuration was selected for the study (Dobi nski & Glazer, 2018;Kasprzak, 2020;Kasprzak et al., 2017;Milsom & Eriksen, 2003).The data collected in the field were then processed in the Res2Dinf program, where average values of ground resistance were calculated, and models of current resistance differentiation were made.

| Meteorological data
Temperature and precipitation data, constituting the meteorological background for the results of ground temperature measurements, were obtained within a year-round measurement cycle from the two meteorological stations (Figure 1c).The first station, Delta OHM HD35AP-S (M1), is a seasonal station located within the raised marine terrace (M1; 18 m a.s.l.), about 380 m SSE from point P5 in the measurement transect.It operates only during the polar expedition, which took place in the summer of 2022.The second station, Hobo U30 (M2), is located near point P2 and is mounted on the measuring cabin at the foot of the slope undergoing solifluction.This study presents the meteorological data, such as temperature, precipitation, and solar radiation, obtained from the M1 station.The measurement series includes data from the period of 19 June to 31 August (see Figure 4).

| Digital ground temperature measurement
In order to compare the results obtained from the optical fibre sensor, the digital thermistor strings based on 2-Wire Bus (TNode, M-Log5W-String, GeoPrecision GmbH) were additionally installed in three places (see Figures 1d and 3a,b).In the gravel beach zone (Z1), a 3 m long string with 15 sensors was installed.On the other hand, on the slope undergoing solifluction (Z2) and within the terrace (Z3), a 2 m long string with 10 sensors was installed.The depth distribution of the sensors in the string is presented in Figure 2a.

| Photonic technologies
A sensing system based on OFDR was used together with innovative, original fibre-optic sensing probes to measure the ground temperature, allowing for continuous recording of this parameter with a high spatial (depth) resolution of 5 cm.
The interrogator had a sensing range of up to 10 m, and due to a special design of the system, the measurement probes could be placed up to 300 m from it; however, the longest of the installed probes had a length of 3 m.The hermetic interrogation unit was supplied with temperature and humidity control (heaters, hygrostat, and fans) that ensured its stable operation independent of external weather The probes were composed of an optical fibre enclosed in a special housing made of materials that had been specifically selected in terms of their thermal conductivity so that they would not disturb the temperature distribution in the ground (Figure 2b-d).The probe was terminated with a tapered end to protect against mechanical damage, facilitating installation in the ground.
The system and sensing probes underwent laboratory tests performed before their installation.These tests were conducted to assess the system's performance for temperature measurement within a range À40 to 60 C. The worst-case scenario measurement accuracy was determined to be ±1.5 C. For obtaining presented measurement results, the system was configured to automatically power up every 2 h and perform measurements for all connected probes.

| Sediments and permafrost
The sediments in the study area have different origins and ages, as determined by geological recognition.The measurement transect includes three morphogenetic and lithological differentiated zones (Figures 1c and 3c): • Zone 1 (Z1)-gravel beach.It is a zone with a width of 120-130 m with two storm ridges: contemporarily transformed by marine processes and old with hinterland.A thicker gravel layer (3-4 m) is deposited on the abrasion platform adjacent to the dead cliff of the 22-30 m a.s.l. raised marine terrace.
• Zone 2 (Z2)-slope undergoing solifluction.It includes a dead cliff.Its inclination is variable: the lower and middle more gently sloping and muddy (clay, loam), varied with solifluction lobs, and upper, more sloping, built of gravel and dry (Repelewska-Pekalowa et al., 2013).
The bedrock consists rocks of the Calypsostranda group (Late Eocene-Early Oligocene), with a series of sandstones, mudstones, shales, coals (NPI, 2016).A list of lithological descriptions of drillings at fibre-optic sensor installation points is presented in Table 1.
The ERT model shows a clear lithological differentiation of the substrate, referring to the separated morphogenetic zones and the extent of the permafrost (Figure 3d,e).The occurrence of permafrost is well visible along the whole transect but particularly within zone 3 (Z3).Very high resistance values occur here from about 2 m from the ground level and constitute a continuous zone up to the edge of the terrace.Within zone 2 (Z2), a thickness of about 2 m of the PAL was found (data from a T A B L E 1 Lithology of sediments at fibre-optic sensor installation points; location of drillings as in Figure 3.

Morphogenetic zone Number of drilling Depth (m) Lithology
Z1-gravel beach P1 0.00-2.602.60-3.00 Gravels and pebbles (from 1.6 m higher share of clasts above 2 cm) with a significant admixture of coarsegrained sands, brownish grey; as above, partially frozen P2 0.00-1.801.80-2.00 Gravels and pebbles with large amounts of sharpedged clasts of stones in the top, brownish grey; as above, partially frozen Z2-slope undergoing solifluction P3 0.00-1.101.10-1.501.50-1.601.60-2.00 Massive silt with remains of malacofauna, dark grey; in the matrix, single poorly rounded gravels; angular coarse-and medium-grained clasts in a silty matrix, dark grey; massive muddy-sandy diamicton with numerous sharp-edged clasts, dark grey (weathered bedrock); as above, partially frozen Z3-raised marine terrace P4 0.00-0.300.30-1.801.80-2.00 Vari-grained sands and gravels with an admixture of organic matter (skeletal tundra soil), dark grey; varigrained sands and gravels with remains of malacofauna; accessory share of angular clasts of rocks, brownish grey; as above, partially frozen P5 0.00-0.20 0.20-1.701.70-2.00 Vari-grained sands and gravels with an admixture of organic matter (skeletal tundra soil), dark grey; varigrained sands and gravels with remains of malacofauna, brownish grey; as above, partially frozen geological survey); however, the resistance pattern was varied, which may also be influenced by lithological diversity.Within zone 1 (Z1), the core with high resistance values is clearly visible.Geological evaluation confirmed the presence of mudstone from a depth of about 2.5 m.The influence of seawater is more evident closer to the coastal zone, as seen in the form of low resistance values.

| Weather conditions
At the Calypsobyen meteorological station (M1), the average daily air temperature at a height of 200 cm above ground level (a.g.l.) for the entire measurement period was 6.2 C (Figures 1 and 4).The highest average daily air temperature, 8.8 C, was recorded on 13 July, while the lowest average of 2.5 C was recorded on 20 June.The maximum air temperature of 9.1 C, was recorded on July 13, and the minimum air temperature of 2.3 C was recorded on June 20.The total precipitation of the entire measurement period was 60.6 mm.There were 22 days with precipitation above 0.1 mm (almost 34% of measurement days), including 9 days with precipitation above 1.0 mm (almost 14% of measurement days).The highest daily precipitation of 16.2 mm occurred on 18 August.It is worth noting that 17 days with rainfall (77.2% of all days with rainfall) occurred in series (i.e., at least 2 days with rainfall in a row).The average daily value of solar radiation at the height of 200 cm a.g.l. in the analysed period was 131.4 W m À2 .The strongest radiation occurred on 8 July and amounted to 319.1 W m À2 , while the lowest average daily solar radiation value of 38 W m À2 was recorded on 18 August.On that day, the second-largest daily rainfall in the period considered was also recorded (Figure 4).

| Ground temperature measurements: First results
Examples of temperature results obtained with the use of thermistors and fibre-optic sensors recorded in the tested transect are shown in Figure 5a-d.It can be seen that the correlation of the results is very good, which is confirmed by the data presented in Table 2.
Analysing the data from Table 2, it can be seen that the averaged absolute values of the differences are not greater than 0.5 C. Significantly lower mean values were obtained when actual (not absolute) differences were averaged.Values around zero indicate that there is no discernible trend in the differences between the two sensor types.
In other words, the differences appear to be random.
Fibre-optic distributed sensors are well known to be sensitive not only to the temperature but also to strain, such as from bending, stresses, and vibrations (Palmieri et al., 2022).If the optical fibre is stressed in an uncontrolled way during the assembly of the probe, it may result in false readings in the context of temperature measurement.An example of such an undesirable occurrence is shown in Figure 5d, where temperature readings from an ill-formed fibre-optic probe are compared to readings from thermistor sensors.It can be seen from this figure that the readings obtained by both sensors are completely different.While the temperature measured with thermistors logically decreases with depth, the graph showing the temperature from fibre-optic sensors shows an inflection that is hard to explain (in terms of temperature changes in the profile).Being aware of this, all probes were calibrated and checked before being installed in the ground, and ill-formed probes were not used in our measurements.

| DISCUSSION
The technology of fibre-optic distributed measurement allows for a significant increase in the spatial resolution of temperature measurement.Our system used a spatial (depth) resolution of 5 cm.It should be stressed that such a resolution was assumed to develop an innovative system that was intended to be validated in the harsh conditions of the Arctic.The resolution of 5 cm was sufficient for this purpose.
Nevertheless, the spatial resolution of measurement systems based on a distributed fibre-optic sensor can be much higher-even in the T A B L E 2 Differences between readings from thermistors and fibre-optic sensors averaged over the vertical profile of investigated positions in the measurement transect.

Morphogenetic zones
The average value of the difference between the thermistors and optical fibre sensors ( C) The average of the difference in absolute values between the thermistors and optical fibre sensors ( C) developed by the authors, involves the construction of a multiparameter-sensing system that will enable the separation of the signals originating from temperature and strain changes.This solution aims to provide a method to distinguish between the effects of temperature and strain on the sensor readings.
The results presented in Figure 5a-d  Achieving higher accuracy and resolution in PAL measurements can provide several benefits.First of all these measurements of higher spatial resolution can reveal more detailed information on the spatial distribution and thickness of PAL, as well as changes in temperature and moisture within the PAL.This can help to better understand the processes driving permafrost thaw and inform predictive models.Such

(
reaching approximately 1.7 m).The trend of increasing PAL thickness has intensified in the last decade.Additionally, there has been an observation of increasingly stronger drying of the ground during this period.The results of PAL thickness measurements from the Calypsostranda site were deemed representative of the areas influenced by the North Atlantic and were included as Site P1 in the international Circumpolar Active Layer Monitoring (CALM) program (Repelewska-Pękalowa & Pękala, 2004).The current measurement transect along which the optical fibre sensors were installed, is located near Calypsobyen, which is approximately 500 m northwest of the Polar Station of Maria Curie-Skłodowska University.It is also situated around 2 km from the mouth of Recherchefjorden.The sensors were installed on the edge zone pacifically on the dead cliff of the raised marine terrace at an elevation of 22-30 m above sea level (Figure1c).

F
Temperature measurement sensors: (a) scheme of the arrangement of the digital thermistor sensors (TNode) in the vertical profile (photo of sensor-Source: GeoPrecision GmbH website); (b) scheme of optical fibre sensor probe; (c) optical fibre sensor probe (Photo InPhoTech); (d) temperature measurement concept of the system, based on the backscattered light.[Colour figure can be viewed at wileyonlinelibrary.com] conditions.The system also used an optical switch for fully automatic switching between optical fibres that connected to each of the sensing probes installed at different locations from a single place.The system was powered with photovoltaic panels and a wind turbine and was equipped with a battery for energy storage.Its power consumption was limited to 90 W. F I G U R E 3 (a) Location of the electrical resistivity tomography (ERT) transect and the installation of sensors system (background: orthophoto map); (b) location of the ERT transect and the installation of sensors system (background: digital elevation model); (c) geological section along the ERT transect; (d) ERT model resistivity with topography-unit electrode spacing = 4 m; (e) ERT model resistivity with topography-unit electrode spacing = 1.25 m; the red arrows indicate the locations of geological drillings and optical fibre sensors installation points.[Colour figure can be viewed at wileyonlinelibrary.com]

4
Course of weather conditions at the meteorological station in Calypsobyen (M1).[Colour figure can be viewed at wileyonlinelibrary.com] order of millimeters.It will only depend on the capabilities of the components (especially the laser) used to build the system and will consequently affect the price of the device.Like any technology, fibre-optic sensors used to measure temperature have their limitations, the most important of which seems a socalled 'cross-sensitivity effect'.Under this term lies the sensitivity of the system to simultaneous changes in temperature and strain.All fibre-optic distributed sensors, except those based on Raman scattering, are subject to this phenomenon(Palmieri et al., 2022).In the context of constructing discussed measurement systems, there are two solutions to this metrological problem.The first solution, as employed by the authors in this work, involves the careful design of the sensor housing to prevent strain transfer to the internal optical fibre.It also includes accurate calibration and meticulous selection of measuring probes before their use in ground measurements, as demonstrated in Figure 5d.The second solution, which is still being F I G U R E 5 Sample results of temperatures in the ground profile.05/08/2022 h 1:00 p.m. form optical fibre sensor comparison with the results of measurements from the digital temperature sensor: (a) P1/S1 installation point-beach, (b) P3/S2 installation point-solifluction slope, (c) P5/S3 installation point-raised marine terrace; (d) example of 'temperature' readings from a fibre-optic compared to temperature readings from digital temperature sensor when optic fibre is under uncontrolled stress.[Colour figure can be viewed at wileyonlinelibrary.com] information cannot be obtained by traditional methods.In addition, accurate and detailed PAL measurements can help to identify areas of high risk for landslides and other hazards associated with permafrost thaw.As permafrost is an important component of the global carbon cycle, the accurate measurements of PAL conditions can help to better understand the role of permafrost in the carbon balance and its response to climate change.It is also worth noting that more precise information about temperature soil conditions monitored in longer periods can lead to better predictions of future ground temperature changes, which is particularly important in the context of global warming and climate change.6| CONCLUSIONSInnovative photonic temperature measuring system, based on OFDR, that was developed and validated in the harsh polar conditions of Spitsbergen has significantly increased the spatial resolution of temperature measurements in the ground profile.Such increased resolution is extremely important in investigating permafrost.Having precise data on the temperature distribution in the substrate enable accurate determination of the thickness of the PAL.It is obvious that this thickness can be different in different years, and the changes can be relatively small.In such cases, a measuring system with low spatial resolution is not fully useful for monitoring such subtle changes.As a consequence, a better measuring system will allow for a better description of the current state of the permafrost and the layers above it in relation to their progressive degradation.In the longer term, it will make possible a better prediction of the pace of changes in the polar environment and will open up previously unavailable opportunities in the field of climate change monitoring and forecasting.ACKNOWLEDGMENTSProject financed by European funds under the European Regional Development Fund and The National Centre for Research and Development: 'Autonomous system of fibre optic quasi-distributed temperature sensor for ground temperature measurement' (SPILOD) no.POIR.04.01.01-00-0031/19-00.
and Table 1 are representative of other measurements performed during the measurement campaign carried out in 2022.The small differences between the readings obtained from the thermistors and fibre-optic sensors indicate the usefulness of this new technology for permafrost measurements.It is important to note that if this technology is effective in such extreme conditions, it can be used for ground surveys in other areas as well.It should be noted that the term 'differences' is being used intentionally instead of 'measurement error' as both methods have uncertainties and neither of them can be considered absolute.