Next Article in Journal
Heat Impact of Urban Sprawl: How the Spatial Composition of Residential Suburbs Impacts Summer Air Temperatures and Thermal Comfort
Previous Article in Journal
Characteristics of Surface Temperature Inversion at the Muztagh-Ata Site on the Pamir Plateau
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Atmosphere
Volume 16
Issue 8
10.3390/atmos16080898
atmosphere-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Polar Mesospheric Winter Echoes Observed with ESRAD in Northern Sweden During 1996–2021

by
Evgenia Belova
1,
Simon Nils Persson
2,
Victoria Barabash
3,* and
Sheila Kirkwood
1
1
Swedish Institute of Space Physics, Box 812, S-981 28 Kiruna, Sweden
2
Department of Computer Science, Electrical and Space Engineering, Luleå University of Technology, S-971 87 Luleå, Sweden
3
Division of Space Technology, Department of Computer Science, Electrical and Space Engineering, Luleå University of Technology, Box 812, S-981 28 Kiruna, Sweden
*
Author to whom correspondence should be addressed.
Atmosphere 2025, 16(8), 898; https://doi.org/10.3390/atmos16080898
Submission received: 26 May 2025 / Revised: 11 July 2025 / Accepted: 20 July 2025 / Published: 23 July 2025
(This article belongs to the Section Upper Atmosphere)
Browse Figures
Versions Notes

Abstract

Polar Mesosphere Winter Echoes (PMWEs) are relatively strong radar echoes from 50–80 km altitudes observed at a broad frequency range, at polar latitudes, mainly during equinox and winter seasons. Most PMWEs can be explained by neutral air turbulence creating structures in the mesosphere and enhanced electron density. We have studied the characteristics of PMWEs and their dependence on solar and geophysical conditions using the ESrange RADar (ESRAD) located in northern Sweden during 1996–2021. We found that PMWEs start in mid-August and finish in late May. The mean daily occurrence rate varied significantly during the PMWE season, showing several relative maxima and a minimum in December. The majority of PMWEs were observed during sunlit hours at 60–75 km. Some echoes were detected at 50–60 km. The echo occurrence rate showed a pronounced maximum near local noon at 64–70 km. During nighttime, PMWEs were observed at about 75 km. PMWEs were observed on 47% of days with disturbed conditions (enhanced solar wind speed, Kp index, solar proton, and X-ray fluxes), and on only 14% of days with quiet conditions. Elevated solar wind speed and Kp index each accounted for 30% of the days with PMWE detections.

1. Introduction

During the last two decades, many experimental studies of Polar Mesosphere Winter Echoes (PMWEs) [1], which are relatively strong radar echoes arising from altitudes between 55 and 80 km during non-summer months [2], have been reported. Most of the PMWEs can be explained by neutral air turbulence that creates coherent structures in the mesosphere (for reviews and references, see [3]). The turbulence has been classified as pure or dusty where the latter involves meteor smoke particles [4,5,6,7]. These charged meteor particles create similar but less pronounced effects within the PMWE layers as do ice particles inside the Polar Mesosphere Summer Echoes (PMSEs) at higher altitudes between 80 and 90 km [8,9,10]. PMWEs were frequently observed when the electron number density in the mesosphere was significantly increased, e.g., [1,11,12,13]. This can be produced both by solar protons and X-rays related to coronal mass ejections and solar flares, respectively, e.g., [2,14,15]. Strengthened electron precipitation from the magnetosphere during high-speed solar wind streams [16,17] has also substantially amplified the PMWE occurrence rate [8]. Belova [18], observing high horizontal speeds in some PMWE events, i.e., structures leading to radar reflection moving in the polar winter mesosphere with horizontal velocities higher than 300 m/s, proposed four mechanisms of such echo generation that are related to microbaroms, i.e., infrasound waves at 0.1–0.35 Hz frequencies produced by ocean swell.
Radar observations of polar mesospheric winter echoes have previously been reported from several locations at high northern latitudes, including Fairbanks, USA, e.g., [19,20]; Kiruna, Sweden, e.g., [1,21]; Andøya and Tromsø, Norway, e.g., [2,12,22,23,24,25]. In Antarctica, PMWEs have been increasingly observed since 2004, e.g., [8,11,26,27,28]. These observations have been made by various radar systems operating with different sensitivities and frequencies in the range between 2 and 224 MHz, although most observations have been made at 53.5 MHz [29] and 52 MHz [1]. Implementation of different radar power systems and sensitivities gave varying limits for echo detection, and thus varying occurrence frequencies and intensities. For example, the more powerful Middle Atmosphere Alomar Radar System (MAARSY) [30] allowed for the detection of turbulence-dominated echoes below 80 km all year round and regularly observed even weak winter echoes [31].
Continuous radar measurements of the polar mesospheric winter echoes covering longer periods are essential for statistical investigation of their morphology and nature. Zeller [22] presented the mean characteristics of PMWEs based on measurements with the ALomar WINd radar (ALWIN) radar at Andøya, Norway, for four winter seasons (2001–2005). The averaged occurrence frequency was estimated to be about 2.9%, with a maximum occurrence height at 70 km for daytime and 77 km for nighttime. Nishiyama [27] reported height and time variations in the PMWE occurrence observed by the Program of the ANtarctic SYowa MST/IS (PANSY) radar from March to September 2013. Their results demonstrated that PMWEs were mainly detected during sunlit hours, which can be explained by photochemistry. They also identified the rapid increases in the daily PMWE occurrence rate associated with solar proton events (SPEs) as well as with geomagnetic disturbances. Kirkwood [8] presented observations of PMWEs at Troll, Antarctica, with a 54 MHz wind-profiler radar, MARA, during two austral winters of 2012–2013. They studied the altitude–seasonal variations in the PMWE occurrence rate and its dependence on solar activity and geomagnetic disturbances. It was shown that the high number of days with PMWEs (60% of all observation days) were mostly related to the effects produced by incoming high-speed solar wind streams (HSSs) to the Earth rather than to solar proton events. Further, Tanaka [32], in a case study, by analysing the satellite and radar data from both the Arctic and Antarctica, suggested that magnetospheric plasma waves excited by the interaction of HSSs with the magnetosphere caused the precipitation of energetic electrons that led to an increase in ionisation at a 55–80 km altitude and the detection of PMWEs. Recently, Latteck [31] presented a comprehensive study of the characteristics of polar mesospheric echoes based on observations with two MST radars at Andøya (69.30° N, 16.04° E), Norway, for two decades. Our paper studies variations in the PMWE occurrence rate based on a long data series from the 52 MHz ESrange RADar (ESRAD) at Esrange (67.9° N, 21.1° E), Sweden, during the period 1996–2021. One of our goals is to examine whether the PMWE occurrence characteristics in terms of altitude, time of day, and season are similar or different at these two sites, Andøya and Esrange, located in northern Scandinavia. We also aim to follow the concept described in [8], and conduct a statistical study on the relation between the PMWE occurrence and solar and geophysical parameters during an extended period.

2. PMWE Observations by ESRAD During 1996–2021

2.1. ESRAD

ESRAD is a mesosphere–stratosphere–troposphere radar located at Esrange, northern Sweden. The purpose of the radar is to provide information on winds, waves, turbulence, and layering from the troposphere up to the mesopause. It was in continuous operation from July 1996 to August 2023, with only minor breaks due to technical problems and upgrades. The radar operated at 52 MHz, with a nominal peak power of 72 kW. However, during the last few years of operation it experienced progressive failure of the power blocks, which led to only 30 kW transmit power being available by 2021. Originally, the ESRAD antenna array consisted of 140 five-element Yagis, with the antenna field extended to 284 in 2004. Whole array was divided into six segments and each of them was connected to one receiver. In 2015, ESRAD was updated with 12 new receivers. The radar transmitted vertically with the whole antenna array, but for reception one can use 6 (12 after December 2015) segments in different combinations. ESRAD ran continuously, cycling between experimental modes dedicated to measurements in the lower troposphere, troposphere-stratosphere, or mesosphere. A typical cycle measures for about 1 min in each mode, repeating every 3–4 min; however, sometimes special cycles, including special experiments, optimised for specific goals, were used. More information about ESRAD can be found in [33].

2.2. ESRAD Data

We looked at the period from 17 December 1996 to 31 January 2021, and four radar modes designed for mesospheric studies: fca_1200, fca_600c, fca_4500, and fca_900. These modes were used during different periods in accordance with the scientific goals and technical setups of the radar. The experiments cover the entire period of interest (with some gaps), and their parameters are shown in Table 1. Some data (especially for 2013–2015) were lost in transferring due to human error.

2.3. PMWE Observations

We used the data of the altitude profiles of the backscattered power where the returns from all receivers were coherently combined. There are two different coherent integration (CI) times: short and long. For detection of PMWEs, we used data with longer CI times of about 1s in order to increase the signal-to-noise ratio (SNR) and, hence, to detect more PMWE events. Further, these power profiles were 1 min averaged.
Then we calculated SNR according to the following equation:
S N R = S   + N − N e s t N e s t
where S and N are signal and noise power, respectively, S + N is received backscattered power from a particular altitude, and Nest is an estimated noise power. Nest is the median received power averaged over 42–48 km altitude, where we do not expect any coherent backscatter [34]. Noise is a sum of cosmic noise and hardware noise.
We defined that a PMWE event occurs when the SNR is above the threshold Th at any height range of 600 m for altitudes 50–78 km for at least 10 min. The values of Th are 0.3, 0.5, or 1.0, depending on whether there are at least 5, between 2 and 4, or only 1 measurement each 10 min. Values were chosen empirically to avoid misidentifying noise as PMWEs. The PMWE occurrence rate of the day was determined as the ratio between the number of samples where PMWE events were detected and the total number of data samples during the day. This is equivalent to the PMWE duration divided by 24 h. Beside the automatic detection of PMWE events based on the threshold, we also made a manual inspection of the entire data set. This allows us to exclude the days where (i) valid data exist only for a part of the day and at some height ranges, (ii) the data contain invalid values, (iii) the enhanced backscatter is due to reasons other than PMWEs, e.g., due to interference, and (iv) PMSEs are detected at higher altitudes and extend down below 78 km. The automatic detection is based on the echo height range and strength, and therefore allows the detection of ‘PMWEs’ even during summer time. Because it is difficult to separate PMSEs extending to the lower heights and the lower-altitude mesospheric echoes which occur in summer time [31], we used manual inspection to discriminate all such summer echoes. Finally, for each day of the year, we counted the ‘sample days’ as the number of days of observations at this date minus the number of skipped days, as described by (i)–(iv), using data for the entire period of our study (24 years). This was further used for calculation of the mean PMWE occurrence rates (see Section 3).

2.4. Seasonal Variations of PMWE Occurrence Rate

Using 24 years of data, we examined how the PMWE occurrence rate varies with the day of year, time of day, and altitude. The results are presented in Figure 1, Figure 2 and Figure 3. Figure 1a shows the mean daily PMWE occurrence rate for the entire data set. For each day of the year, the occurrence rate was calculated as the total duration of the PMWE events on this day divided by 24 h (the ratio between the number of samples with PMWE events and the number of data samples), and then averaged over the sample days (shown as the red curve); finally, a 6-day running mean was applied. For the majority of dates, more than 12 sample days are available, with the obvious exception for the leap day of February 29. The PMWE occurrence rate shows strong seasonal variability. The PMWE season starts in mid-August and ends in late May. The PMWE occurrence rate has several maxima during this period. The largest rate of about 13% is reached very quickly in the beginning of October; then, the occurrence rate fluctuates at a relatively high level until late December, when it drops down to smaller values. It shows higher values again in March and April.
Figure 1b shows the daily PMWE occurrence rates for each year from 1997 to 2021, smoothed by a 6-day running averaging. Note that there are large data gaps in 2013–2015. We can observe high year-to-year variability in the PMWE occurrence rates, with several strong intensifications in the individual years, which can account for the peaks in the mean PMWE occurrence rate seen in Figure 1a. For example, there are enhanced rates in (1) 2001, 2004–2006, and 2012; (2) April 2001, 2008, 2010, and 2012; (3) October 2001, 2003–2005, 2008, 2016, and 2017; (4) November 2004 and 2016.
In Figure 2, the mean seasonal and local time variations in the PMWE occurrence rate are presented. For each day and each 30 min interval of local time, the occurrence rate was calculated as the ratio of the number of PMWE samples and the total number of data samples. Then these occurrence rates were averaged over 24 years of observations. The red lines indicate the time corresponding to the 98° solar zenith angle, which separates day- and nighttime conditions at a 60 km altitude. It is clearly seen that a majority of winter echoes were detected when the altitudes where the echoes occur were illuminated by the sun. It is interesting to notice that, in the morning hours, PMWEs have been mainly observed approximately 1 h after sunrise.
In Figure 3, the altitude distribution of the mean PMWE occurrence rate is shown as a function of the day of the year. For each altitude range of 600 m and each day of year, occurrence rate is calculated as the ratio between the number of the samples during which PMWEs were detected and the total number of the samples during the day. Then the mean day altitude occurrence rate for a specific date and specific height is calculated by averaging for 24 years. One can notice that the PMWE preferred altitudes are 60–75 km; however, sometimes echoes can occur at as low as 53 km. In December and September, and less obviously in April-May, the lowest PMWEs were observed at somewhat higher altitudes than in the other months.

2.5. Altitude–Diurnal Variation of PMWE Occurrence Rate

The altitude–diurnal distribution of the mean PMWE occurrence rate is presented in Figure 4. For each 600 m height range and 30 min time bin, the occurrence rate was calculated as a ratio of the number of PMWE samplings and the number of all samplings that were made, and then averaged over the entire data set. There is a clear variation in winter echoes: from about 6 to 17 local time, it was possible to observe PMWEs below 60 km; however, in the evening and morning hours, the echo lowest altitude increased, and no echoes were detected below 75 km during nighttime. The highest occurrence rate values were observed around local noon at a 63–72 km altitude.

2.6. Solar and Geophysical Parameters

2.6.1. Data Overview

In Figure 1 we can observe that the PMWE occurrences at certain dates and periods are 2–3 times higher than at other ones. However, it is not possible to find conditions facilitating high occurrence rates when averaging over 24 years. In order to make such a study, we looked at PMWE detections on a daily basis together with solar and geophysical conditions that may affect echo generation. In previous studies, e.g., [1,10,14,17], PMWEs have often been observed during strong geomagnetic disturbances, solar proton events, solar flares (strong X-ray fluxes), and in connection to the arrival of high-speed solar wind streams (HSSs). We choose four parameters, the Kp index, X-ray flux, solar wind speed, and solar proton flux, which can be used as a measure of disturbed conditions and are related to the ionisation at PMWE altitudes. The Kp index is a planetary geomagnetic field index characterising the general level of geomagnetic disturbances, and high Kp values are an indication of geomagnetic substorms/storms during which energetic particles precipitate into the ionosphere. Solar X-rays and solar protons directly produce ionisation at PMWE altitudes. Arriving at Earth, HSSs cause several processes which lead to high energy electron precipitation from the magnetosphere to the ionosphere [16]. Solar and geomagnetic data were taken from the NASA web service OMNI Web Plus (https://omniweb.gsfc.nasa.gov/, accessed on 25 April 2025) [35]. It includes data from 1963 up to the present from satellite- and ground-based instrumentation all around the world hosted by many countries.
In Figure 5, the monthly numbers of days with PMWEs (i.e., days when SNR is equal or exceeds the threshold for minimum 10 min at any height of interest) and sample days are presented. We see that the number of sample days was quite variable during first 6–7 years of ESRAD operation. Moreover, starting from winter 2017–2018, the number of days with PMWEs was rather small, which may be an indication of the diminished sensitivity of ESRAD due to the reduced transmit power. It also turned out that the OMNI data have relatively large gaps in 2000–2003. Taking all this into account, we chose for further study the period from 1 January 2004 to 31 December 2011, when we assume ESRAD had approximately the same PMWE detection sensitivity between the two updates and when the number of monthly observation days was greater than 15 (see Figure 5).

2.6.2. Dependence on Solar and Geophysical Parameters

Figure 6 shows the proportion of PMWE days in all sample days as a function of the Kp index, X-ray flux, solar wind speed, and solar proton flux. The solar parameter data have a time resolution of one hour and the Kp index is estimated for every three hours. We chose the maximum values which occurred on each day to represent the day. The proportion of PMWE days was calculated as a ratio of the number of days when PMWEs were observed and the number of sample days for given maximum values of solar and geophysical parameters observed during these days. We show here only the bins for the parameter values with the number of sample days above nine.
We see that there is no visible effect of solar X-ray flux on the appearance of PMWEs: the proportion of days with echoes remains at about 20% independently of the strength of the X-ray flux. The situation is similar for solar protons until the very high values of the proton fluxes, when the proportion of the PMWE days reaches 50–60%. The occurrence rate grows gradually to its maximum of 60% with increasing Kp values up to 4, and then remains at this level or slightly decreases for the higher Kp values. The solar wind speed has the strongest relation with the PMWE occurrence rate. The higher the speed, the more frequent winter echoes were observed. When solar wind speed exceeds 700 km/s, the occurrence rate reaches its highest value of 75%. Following Kirkwood et al. [8], we have studied this further and considered different combinations of the solar and geophysical parameters in relation to the winter echo occurrence rate. The results are presented in Table 2. The total number of days when the parameters in any combinations were above the thresholds is 1072, including 503 days when PMWEs were detected. Thus, PMWEs appear on 47% of all days with disturbed conditions, while the occurrence at ‘‘quite’’ geophysical and solar conditions is only 14%. The effects of enhanced solar wind speeds and Kp index on the PMWE occurrence rate are about the same (~30%), which is much higher than for the X-rays. The poor statistics on days with only enhanced protons does not allow us to draw conclusions about their influence on PMWEs. When all four parameters were elevated, PMWEs were observed on 92% of such days.

3. Discussion

Latteck [31] presented a study of mesospheric echoes at 69° N at Andøya, Norway, for the period 1999–2019 using successive observations with two MST radars, ALWIN and MAARSY. The authors sorted all observed echoes as PMSEs and lower-altitude mesospheric echoes (LAMEs). The latter were observed during the whole year; however, for non-summer seasons, they were equivalent to PMWEs. In this paper we presented the seasonal, altitude, and diurnal variations in the PMWE occurrence rate based on 24 years of observations with the ESRAD radar, located 250 km southeast of MAARSY. Our results, presented in Section 2, show that the behaviour of the winter echo occurrence rate at ESRAD over the course of the season, with altitude and with time of the day, is similar to that at MAARSY. The majority of the echoes occur during sunlit hours, when the sun illuminates the PMWE altitudes (Figure 2), with an increase in the occurrence rate around the local noon. This indicates a strong relation with the photo-ionisation of the ionospheric D region. Both radars show the fast growth of the PMWE occurrence rate from August to September, with a maximum in late September, a minimum in December-January, an increase again in March-April, and then a more gradual decrease in May (Figure 1). Latteck [31] suggested that these PMWE occurrence rate variations are likely related to seasonal–altitude variations in neutral turbulence produced by gravity waves. The simulations by Hoffmann [36] showed that, from May to September, the high values of the mean kinetic energy of the gravity waves, which is a main source for turbulence, occurred at altitudes above 80 km, and the absence of turbulence below 75 km. In October, the turbulence area moves to lower altitudes and remains there until March-April. The bottom boundary of the PMWE height range goes up in December so that the width of the height range is minimised (Figure 3). This minimum and the one in daily occurrence rate can be related to the shortest sunlit time, during which free electrons needed for the radar backscatter are produced. ALWIN, MAARSY, and ESRAD show similar diurnal–altitude variation in the PMWE occurrence rate. During nighttime, most echoes occur at heights over 75 km; with sunlight, they appear at the lower heights and a show peak in the occurrence rate at 65–70 km from 9 local time (LT) to 14 LT (Figure 6) [22,31]. MAARSY detected echoes at lower altitudes compared to ALWIN and ESRAD at nighttime due to its higher sensitivity, which will be discussed further. Kirkwood [37] explained the day–night differences by the reduction in electron density and increasing electron diffusivity, which result from the attachment of electrons to form negative ions below ∼75 km at night. Additionally, meteoric smoke particles being mostly neutral during sunlit hours, because of a high rate of photo-detachment, scavenge electrons during night [38].
It is interesting to notice the similarity in the intensifications of the mean daily occurrence rates for ESRAD and MAARSY. In October, November, mid-December, March, and April, both radars show increases in the occurrence rate which lasted for a few days. Correlated intensifications may indicate the same reason behind them, e.g., extra ionisation in the mesosphere above both radar sites due to a solar proton event and/or energetic electron precipitation. ESRAD and MAARSY are located 250 km apart, at latitudes 67.9° N and 69.3° N, respectively. Solar proton precipitation is typically extended over a wide latitude range and can be a source of ionisation over both radar sites. Precipitation of high energy electrons, E > 30 keV, which are associated with HSSs, and are able to reach and produce ionisation at altitudes between 50 and 90 km, occur in latitude bands between about 65° and 80° N [16]. Thus, one might expect that, with the arrival of HSSs at Earth, the mesosphere above both radars will be exposed to the energetic electron precipitation, causing the enhancement of the electron density and, hence, the detectability of PMWEs.
The radars have very different sensitivities, i.e., minimum detectable echo strengths, which are determined largely by the product of the transmitted power and the effective area of the receiving antennas. The transmit power and effective area are 36 kW and 1690 m2, respectively, for ALWIN, 72 kW and 3740 m2 for ESRAD after 2004, and 736 kW and 5590 m2 for MAARSY [39]. Differences in the experiment configurations (e.g., the number of coherent integrations) also plays a role. The different sensitivities of the radars can explain why the PMWE occurrence rates for ESRAD are higher than for ALWIN and lower than for MAARSY, as seen from Figure 1, Figure 2 and Figure 3 in this paper and Figure 3 in [31]. We should also notice that different PMWE event detection criteria were applied for ALWIN/MAARSY and ESRAD. Latteck [31] defined a PMWE event as an echo reflectivity exceeding 10−18 m−1 for a minimum duration of 20 min, while, in this paper, we require an SNR exceeding 0.3–1 for at least 10 min.
Using one year of observations by the PANSY radar at the Antarctic Syowa station, Nishiyama [27] found a delay of 1–2 h in the PMWE onset time relative to sunrise. Such behaviour is clearly seen in our results as well (Figure 2). This delay is likely related to stratospheric ozone shadowing sunlight of UV wavelengths [40], which are important for photochemical processes, leading to the detachment of electrons from the negative ions accumulated in the ionospheric D region during nighttime. When the sun is just rising above the horizon (solar zenith angle is 98° for 60 km height), the solar light goes through a column of high ozone concentration at 15–40 km, and the UV flux will be absorbed there before reaching the D region. Later, when the sun has risen sufficiently high and directly illuminates the D region, a sufficient number of electrons are produced due to the ionisation of neutrals, and photo-detachment from the negative ions [41] and meteoric smoke particles [38,42], which allows PMWEs to become detectable by radar.
As mentioned before, previous studies have reported that the PMWE events were often observed during disturbed geomagnetic conditions or solar proton events, e.g., [1,10,18,32]. We confirmed this statistically using PMWE data for eight years. We chose four parameters characterising geomagnetic and solar conditions, and demonstrated that the number of PMWE days increases for high values of solar proton flux and, more obviously, for a high solar wind speed and Kp index. For the last two parameters, we confirmed the results of [8] based on data from the MARA radar for two winter seasons in Antarctica. Latteck and Strelnikova [2] reported a correlation coefficient of 0.35 between the daily PMWE occurrence rate and the maximum values of Kp based on observations with MAARSY during two winters of 2011–2013. In this study, the duration of PMWEs for 24 h was used for the calculation of the occurrence rate together with the 3 h Kp index. Zeller [22] found correlations between the monthly PMWE occurrence rate and solar proton fluxes, and an Ap index of 0.83 and 0.4, respectively, using the ALWIN data for 2001–2005. Ap is a 24 h magnetic index related to the daily maximum of Kp and, hence, it has weak relation to the duration of PMWEs, i.e., to the PMWE occurrence rate. This, together with monthly averaging, might explain the small value of the correlation between the Ap index and PMWE occurrence rate. Since solar proton events can last for several days, the correlation with solar proton fluxes is higher.
Our further analysis of the days with PMWEs and the days with different combinations of geomagnetic and solar disturbances showed the following. Chances to observe PMWEs on the days with Kp > 2 or on the days with a solar wind speed > 450 km s−1 are two times higher compared to the days with undisturbed conditions (see Table 2). This is qualitatively consistent with the results by [8] obtained with MARA at Troll, Antarctica, for 2012–2014. There, PMWEs were reported to be observed on 76% of all days with high solar wind speed and on 60% of days with enhanced Kp, compared to 44% of days with quiet conditions. However, our results are based on better statistics, with 1599 observation days compared to 453 days in [8]. In our study, we found six observation days satisfying the criteria described in Section 2.3, during which the solar proton fluxes were enhanced (while three other parameters were on undisturbed levels). No PMWEs were observed on these dates. On the other hand, Nishiyama [27] was able to relate only two out of six solar proton events to the PMWE occurrence rate.
The majority of PMWEs can be attributed to Bragg scattering of the radar waves on the coherent structures in the electron density caused by neutral air turbulence in the presence (or absence) of charged meteoric smoke particles, e.g., [43]. The intensity of such coherent echoes and, hence, the possibility to detect them, is determined by the background electron density, turbulent dissipation rate, and, to a lesser extent, by the Schmidt number, i.e., the size and concentration of charged dust, e.g., [44,45]. Thus, we can speculate that a low level of turbulence might hypothetically account for the absence of PMWEs on 53% of days with disturbed conditions (Table 2). Then, the electron density was not high enough to compensate for the low turbulence and make the backscatter signal detectable by the radar. On the other hand, following [21], if we consider the non-turbulent hypothesis of the PMWE origin, which involves infrasound waves, then the lack of PMWEs may be an indication of the low amplitudes of these waves in the North Atlantic source region and/or a lack of suitable conditions for their propagation to the mesosphere over the radar sites [18].
We also found that PMWEs were observed on 14% of quiet days, i.e., when all four parameters were at undisturbed levels. This might be due to the following. Solar X-rays and protons have, unlike HSSs, a direct effect on the ionisation of the D region, which ceases immediately after the end of events. Kirkwood [8] showed that PMWEs were observed up to 15 days after the HSS arrival at the Earth, while solar wind speed and the Kp index fall to an undisturbed pre-HSS level within 5–10 days. The authors explained this by high-energy electron precipitation from the radiation belts, which is responsible for ionisation at the PMWE heights. It starts after the HSS onset, when the solar wind speed increases sharply from below 400 to above 450 km s−1, and continues for several days afterwards [16].

4. Conclusions

In this paper, we investigated the occurrence frequency of Polar Mesosphere Winter Echoes and its dependence on solar and geophysical conditions using observations with the atmospheric radar ESRAD located near Kiruna, northern Sweden, during 1996–2021.
We found that the mean daily occurrence varies significantly during the PMWE season. The PMWE season starts in mid-August and finishes in late May. The occurrence rate shows several maxima during the season and a dip in December. The majority of PMWEs in the course of the season were observed during sunlit hours. The echoes were shown to occur mainly at a 60–75 km altitude; however, some echoes were detected at 50 km, and the PMWE altitude range in December was reduced compared to other months. The echo occurrence rate shows a pronounced maximum near local noon at a 64–70 km altitude. During nighttime, PMWEs were observed at higher altitudes of about 75 km. We found that the seasonal, diurnal, and altitude PMWE behaviour was similar to that of PMWEs observed with the ALWIN/MAARSY radars located at Andøya, 250 km northwest of ESRAD.
We also found that the number of days with PMWEs increases with the growth of the solar wind speed and geomagnetic Kp index. Analysis of 2004–2011 showed that PMWEs were observed on 47% of all days with disturbed geophysical and solar conditions, characterised by enhanced solar wind speed, Kp index, solar proton, and X-ray fluxes, and only on 14% of days with quiet conditions. It was found that an elevated solar wind speed and Kp index each accounted for 30% of the days with PMWE detections. This qualitatively confirms the results of [8] for the MARA radar in Antarctica.

Author Contributions

E.B. and V.B. proposed the idea of the paper. E.B. formulated the problems and tasks, and checked the results obtained. S.K. wrote software for the on-line analysis of ESRAD data and the initial version of the software for off-line automatic search for PMWEs. S.N.P. performed the analysis of solar wind and Kp index data, developed further software for PMWE analysis, and produced the figures. These tasks were performed under the supervision of V.B. as a part of S.N.P.’s Master’s thesis project. All authors participated in the discussions of the results. E.B. wrote the manuscript, with substantial contributions from V.B. and S.K. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Radar data requests should be directed to victoria.barabash@ltu.se at Luleå University of Technology, or to evgenia.beloiva@irf.se at the Swedish Institute of Space Physics.

Acknowledgments

ESRAD operation and maintenance was provided by the Esrange Space Center, Swedish Space Corporation (SSC).

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Kirkwood, S.; Barabash, V.; Belova, E.; Nilsson, H.; Rao, N.; Stebel, K.; Osepian, A.; Chilson, P.B. Polar mesosphere winter echoes during solar proton events. Adv. Polar Up. Atmos. Res. 2002, 16, 111–125. [Google Scholar]
  2. Latteck, R.; Strelnikova, I. Extended observations of polar mesosphere winter echoes over Andøya (69°N) using MAARSY. J. Geophys. Res. Atmos. 2015, 120, 8216–8226. [Google Scholar] [CrossRef]
  3. Strelnikov BStaszak, T.; Latteck, R.; Renkwitz, T.; Strelnikova, I.; Lübken, F.J.; Baumgarten, G.; Fiedler, J.; Chau, J.L.; Stude, J.; Rapp, M.; et al. Sounding rocket project “PMWE” for investigation of polar mesosphere winter echoes. J. Atmos. Sol.-Terr. Phys. 2021, 218, 105596. [Google Scholar] [CrossRef]
  4. Rosinski, J.; Snow, R.H. Secondary Particulate Matter from Meteor Vapors. J. Atmos. Sci. 1961, 18, 736–745. [Google Scholar] [CrossRef]
  5. Hughes, D.W. Meteors and cosmic dust. Endeavour 1997, 21, 31–35. [Google Scholar] [CrossRef]
  6. Janches, D.; Heinselman, C.J.; Chau, J.L.; Chandran, A.; Woodman, R. Modeling the global micrometeor input function in the upper atmosphere observed by high power and large aperture radars. J. Geophys. Res. 2006, 111, A07317. [Google Scholar] [CrossRef]
  7. Mathews, J.D.; Briczinski, S.J.; Meisel, D.D.; Heinselman, C.J. Radio and meteor science outcomes from comparisons of meteor radar observations at AMISR Poker Flat, Søndrestrøm, and Arecibo. Earth Moon Planets 2008, 102, 365–372. [Google Scholar] [CrossRef]
  8. Kirkwood, S.; Osepian, A.; Belova, E.; Lee, Y.S. High-speed solar wind streams and polar mesosphere winter echoes at Troll, Antarctica. Ann. Geophys. 2015, 33, 609–622. [Google Scholar] [CrossRef]
  9. Singer, W.; von Zahn, U.; Weiß, J. Diurnal and annual variations of meteor rates at the arctic circle. Atmos. Chem. Phys. 2004, 4, 1355–1363. [Google Scholar] [CrossRef]
  10. Stude, J.; Aufmhoff, H.; Schlager, H.; Rapp, M.; Baumann, C.; Arnold, F.; Strelnikov, B. Measurement report: Rocket-borne measurements of large ions in the mesosphere and lower thermosphere—Detection of meteor smoke particles. Atmos. Chem. Phys. 2025, 25, 383–396. [Google Scholar] [CrossRef]
  11. Nishiyama, T.; Sato, K.; Nakamura, T.; Tsutsumi, M.; Sato, T.; Tanaka, Y.M.; Nishimura, K.; Tomikawa, Y.; Kohma, M. Simultaneous observations of polar mesosphere winter echoes and cosmic noise absorptions in a common volume by the PANSY radar (69.0°S, 39.6°E). J. Geophys. Res. Space Phys. 2018, 123, 5019–5032. [Google Scholar] [CrossRef]
  12. Lübken, F.J.; Singer, W.; Latteck, R.; Strelnikova, I. Radar measurements of turbulence, electron densities, and absolute reflectivities during polar mesosphere winter echoes (PMWE). Adv. Space Res. 2007, 40, 758–764. [Google Scholar] [CrossRef]
  13. Havnes, O.; Kassa, M. On the sizes and observable effects of dust particles in polar mesospheric winter echoes. J. Geophys. Res. Atmos. 2009, 114, D09209. [Google Scholar] [CrossRef]
  14. Belova, E.; Kawnine, M.; Häggström, I.; Sergienko, T.; Kirkwood, S.; Tjulin, A. Tristatic observation of polar mesosphere winter echoes with the EISCAT VHF radar on 8 January 2014: A case study. Earth Planets Space 2018, 70, 110. [Google Scholar] [CrossRef] [PubMed]
  15. Brattli, A.; Blix, T.A.; Lie-Svendsen, O.; Hoppe, U.P.; Lübken, F.J.; Rapp, M.; Singer, W.; Latteck, R.; Friedrich, M. Rocket measurements of positive ions during polar mesosphere winter echo conditions. Atmos. Chem. Phys. 2006, 6, 5515–5524. [Google Scholar] [CrossRef]
  16. Meredith, N.P.; Horne, R.B.; Lam, M.M.; Denton, M.H.; Borovsky, J.E.; Green, J.C. Energetic electron precipitation during high-speed solar wind stream driven storms. J. Geophys. Res. 2011, 116, A05223. [Google Scholar] [CrossRef]
  17. Tsurutani, B.T.; Hajra, R.; Echer, E.; Lakhina, G.S. Comment on “First Observation of Mesosphere Response to the Solar Wind High-Speed Streams” by W. Yi et al. J. Geophys. Res. Space Phys. 2019, 124, 8165–8168. [Google Scholar] [CrossRef]
  18. Belova, E.; Barabash, V.; Godin, O.A.; Kero, J.; Näsholm, S.P.; Vorobeva, E.; Le Pichon, A. High-speed Echoes in the Polar Winter Mesosphere: Infrasound as a Probable Cause. Adv. Space Res. 2023, 72, 3181–3201. [Google Scholar] [CrossRef]
  19. Ecklund, W.L.; Balsley, B.B. Long-term observations of the arctic mesosphere with the MST radar at Poker Flat, Alaska. J. Geophys. Res. 1981, 86, 7775–7780. [Google Scholar] [CrossRef]
  20. Balsley, B.B.; Ecklund, W.L.; Fritts, D.C. VHF echoes from the high-latitude mesosphere and lower thermosphere: Observations and interpretations. J. Atmos. Sci. 1983, 40, 2451–2466. [Google Scholar] [CrossRef]
  21. Kirkwood, S.; Chilson, P.; Belova, E.; Dalin, P.; Häggström, I.; Rietveld, M.; Singer, W. Infrasound—The cause of strong Polar Mesosphere Winter Echoes? Ann. Geophys. 2006, 24, 475–491. [Google Scholar] [CrossRef]
  22. Zeller, O.; Zecha, M.; Bremer, J.; Latteck, R.; Singer, W. Mean characteristics of mesosphere winter echoes at mid- and high latitudes. J. Atmos. Sol. Terr. Phys. 2006, 68, 1087–1104. [Google Scholar] [CrossRef]
  23. Renkwitz, T.; Latteck, R.; Strelnikova, I.; Johnsen, M.G.; Chau, J.L. Characterization of polar mesospheric VHF radar echoes during solar minimum winter 2019/2020. Part I: Ionization. J. Atmos. Sol.-Terr. Phys. 2021, 221, 105684. [Google Scholar] [CrossRef]
  24. Hall, C.M.; Manson, A.H.; Meek, C.M.; Nozawa, S. Isolated lower mesospheric echoes seen by medium frequency radar at 70° N, 19° E. Atmos. Chem. Phys. 2006, 6, 5307–5314. [Google Scholar] [CrossRef]
  25. La Hoz, C.; Havnes, O. Artificial modification of polar mesospheric winter echoes with an RF heater: Do charged dust particles play an active role? J. Geophys. Res. 2008, 113, D19205. [Google Scholar] [CrossRef]
  26. Morris, R.J.; Klekociuk, A.R.; Holdsworth, D.A. First observations of Southern Hemisphere polar mesosphere winter echoes including conjugate occurrences at ≈ 69°S latitude. Geophys. Res. Lett. 2011, 38. [Google Scholar] [CrossRef]
  27. Nishiyama, T.; Sato, K.; Nakamura, T.; Tsutsumi, M.; Sato, T.; Kohma, M.; Nishimura, K.; Tomikawa, Y.; Ejiri, M.K.; Tsuda, T.T. Height and time characteristics of seasonal and diurnal variations in PMWE based on 1-year observations by the PANSY radar (69.0°S, 39.6°E). Geophys. Res. Lett. 2015, 42, 2100–2108. [Google Scholar] [CrossRef]
  28. Kohma, M.; Sato, K.; Nishimura, K.; Tsutsumi, M.; Sato, T. A statistical analysis of the energy dissipation rate estimated from the PMWE spectral width in the Antarctic. J. Geophys. Res.: Atmos. 2020, 125, e2020JD032745. [Google Scholar] [CrossRef]
  29. Latteck, R.; Bremer, J. Long-term variations of polar mesospheric summer echoes observed at Andøya (69°N). J. Atmos. Sol. Terr. Phys. 2017, 163, 31–37. [Google Scholar] [CrossRef]
  30. Latteck, R.; Singer, W.; Rapp, M.; Vandepeer, B.; Renkwitz, T.; Zecha, M.; Stober, G. MAARSY—The new MST radar on Andøya: System description and first results. Radio Sci. 2012, 47, RS1006. [Google Scholar] [CrossRef]
  31. Latteck, R.; Renkwitz, T.; Chau, J.L. Two decades of long-term observations of polar mesospheric echoes at 69°N. J. Atmos. Sol.-Terr. Phys. 2021, 216, 105576. [Google Scholar] [CrossRef]
  32. Tanaka, Y.M.; Nishiyama, T.; Kadokura, A.; Ozaki, M.; Miyoshi, Y.; Shiokawa, K.; Oyama, S.L.; Kataoka, R.; Tsutsumi, M.; Nishimura, K.; et al. Direct comparison between magnetospheric plasma waves and polar mesosphere winter echoes in both hemispheres. J. Geophys. Res. Space Phys. 2019, 124, 9626–9639. [Google Scholar] [CrossRef]
  33. Chilson, P.B.; Kirkwood, S.; Nilsson, A. The Esrange MST radar: A brief introduction and procedure for range validation using balloons. Radio Sci. 1999, 34, 427–436. [Google Scholar] [CrossRef]
  34. Kirkwood, S.; Belova, E.; Satheesan, K.; Narayana Rao, T.; Rajendra Prasad, T.; Satheesh Kumar, S. Fresnel scatter revisited—comparison of 50 MHz radar and radiosondes in the Arctic, the Tropics and Antarctica. Ann. Geophys. 2010, 28, 1993–2005. [Google Scholar] [CrossRef]
  35. NASA. Goddard Space Flight Center, Space Physics Data Facility, OMNI Web Plus. Available online: https://omniweb.gsfc.nasa.gov/ (accessed on 7 April 2025).
  36. Hoffmann, P.; Becker, E.; Singer, W.; Placke, M. Seasonal variation of mesospheric waves at northern middle and high latitudes. J. Atmos. Sol. Terr. Phys. 2010, 72, 1068–1079. [Google Scholar] [CrossRef]
  37. Kirkwood, S. Polar mesosphere winter echoes—A review of recent results. Adv. Space Res. 2007, 40, 751–757. [Google Scholar] [CrossRef]
  38. Rapp, M. Charging of mesospheric aerosol particles: The role of photodetachment and photoionization from meteoric smoke and ice particles. Ann. Geophys. 2009, 27, 2417–2422. [Google Scholar] [CrossRef]
  39. Latteck, R.; Murphy, D.J. Climatological comparison of polar mesosphere summer echoes over the Arctic and Antarctica at 69°. Ann. Geophys. 2024, 42, 55–68. [Google Scholar] [CrossRef]
  40. Macotela, E.L.; Clilverd, M.A.; Manninen, J.; Thomson, N.R.; Newnham, D.A.; Raita, T. The effect of ozone shadowing on the D-region ionosphere during sunrise. J. Geophys. Res. Space Phys. 2019, 124, 3729–3742. [Google Scholar] [CrossRef]
  41. Turco, R.P.; Sechrist, C.F., Jr. An investigation of the ionospheric D region at sunrise: 3, time variations of negative-ion and electron densities. Radio Sci. 1972, 7, 725–737. [Google Scholar] [CrossRef]
  42. Havnes, O.; La Hoz, C.; Rietveld, M.T.; Kassa, M.; Baroni, G.; Biebricher, A. Dust charging and density conditions deduced from observations of PMWE modulated by artificial electron heating. J. Geophys. Res. 2011, 116, D24203. [Google Scholar] [CrossRef]
  43. Staszak, T.; Strelnikov, B.; Latteck, R.; Renkwitz, T.; Friedrich, M.; Baumgarten, G.; Lübken, F.J. Turbulence generated small-scale structures as PMWE formation mechanism: Results from a rocket campaign. J. Atmos. Sol.-Terr. Phys. 2021, 217, 105559. [Google Scholar] [CrossRef]
  44. Lübken, F.J.; Strelnikov, B.; Rapp, M.; Singer, W.; Latteck, R.; Brattli, A.; Hoppe, U.P.; Friedrich, M. The thermal and dynamical state of the atmosphere during polar mesosphere winter echoes. Atmos. Chem. Phys. 2006, 6, 13–24. [Google Scholar] [CrossRef]
  45. Fritts, D.C.; Alexander, J. Gravity wave dynamics and effects in the middle atmosphere. Rev. Geophys. 2003, 41, 1003. [Google Scholar] [CrossRef]
Atmosphere 16 00898 g001
Figure 1. (a) Averaged day-to-day variation in PMWE occurrence rate smoothed by a 6-day running mean. The red curve shows the number of sample days used in the averaging. The data are for the period between 17 December 1996 and 31 January 2021. (b) The same as for (a) but for each year from 1997 to 2021.
Figure 1. (a) Averaged day-to-day variation in PMWE occurrence rate smoothed by a 6-day running mean. The red curve shows the number of sample days used in the averaging. The data are for the period between 17 December 1996 and 31 January 2021. (b) The same as for (a) but for each year from 1997 to 2021.
Atmosphere 16 00898 g001
Atmosphere 16 00898 g002
Figure 2. The mean seasonal and local time variations in PMWE occurrence rate. The red lines indicate solar zenith angle of 98° separating daytime from nighttime conditions at 60 km altitude.
Figure 2. The mean seasonal and local time variations in PMWE occurrence rate. The red lines indicate solar zenith angle of 98° separating daytime from nighttime conditions at 60 km altitude.
Atmosphere 16 00898 g002
Atmosphere 16 00898 g003
Figure 3. The mean seasonal and altitude variation in PMWE occurrence rate.
Figure 3. The mean seasonal and altitude variation in PMWE occurrence rate.
Atmosphere 16 00898 g003
Atmosphere 16 00898 g004
Figure 4. The diurnal–altitude distribution of PMWE occurrence rate averaged for period between 17 December 1996 and 31 January 2021.
Figure 4. The diurnal–altitude distribution of PMWE occurrence rate averaged for period between 17 December 1996 and 31 January 2021.
Atmosphere 16 00898 g004
Atmosphere 16 00898 g005
Figure 5. (a) The monthly numbers of sample days are shown as a red curve. The grey bars indicate no available data; (b) the percentage of PMWE days (blue bars) in the sample days each month.
Figure 5. (a) The monthly numbers of sample days are shown as a red curve. The grey bars indicate no available data; (b) the percentage of PMWE days (blue bars) in the sample days each month.
Atmosphere 16 00898 g005
Atmosphere 16 00898 g006
Figure 6. Proportion of PMWE days out of the sample days for period between 1 January 2004 and 31 December 2011 as a function of daily maximum value of (a) Kp index, (b) X-ray flux, (c) solar wind speed, and (d) solar proton flux. Only observations for solar zenith angle less than 98° (i.e., during daylight) are considered for X-rays.
Figure 6. Proportion of PMWE days out of the sample days for period between 1 January 2004 and 31 December 2011 as a function of daily maximum value of (a) Kp index, (b) X-ray flux, (c) solar wind speed, and (d) solar proton flux. Only observations for solar zenith angle less than 98° (i.e., during daylight) are considered for X-rays.
Atmosphere 16 00898 g006
Table 1. The parameters of the ESRAD modes used in the paper.
Table 1. The parameters of the ESRAD modes used in the paper.
Modefca_1200fca_600cfca_4500fca_900
Period of usage17 December 1996—2 May 1997 20 May 1997–29 August 199920 October 1999—16 December 201525 March 2016—31 January 2021
Pulse repetition frequency, Hz10241024/130013001300
Codenone8-bit complementary8-bit complementarynone
Height resolution, m600600600600
Start height, m5000480048001050
Stop height, m99,20099,60099,600100,650
Number of coherent integrations (long)10245126401280
Table 2. Distribution of days with PMWEs and sample days for the period between 1 January 2004 and 31 December 2011 (June, July, and August were excluded) as a function of the solar and geophysical disturbance level each day represented by Kp index, solar wind speed, solar proton flux, and solar X-ray flux. Each day is categorised into 1 of 16 possible states according to whether the threshold value for each parameter is exceeded at any time during the day. Thresholds are Kp = 2, Vsw = 450 km/s, proton flux = 1 cm−2s−1sr−1 (for >10 MeV protons), and X-ray flux = 5 × 10−7 Wm−2 (for 0.5–4.0 Å X-rays, only daytime is considered). Ones indicate that the threshold for the corresponding parameter in columns 1–4 was exceeded. Thresholds are the same as in [8]. Column 5 shows the number of sample days in each category, and column 6 shows the number of days when PMWEs were observed. Column 7 shows the percentage of sample days when PMWEs were observed. These are shown in bold type when the number of sample days is greater than 10.
Table 2. Distribution of days with PMWEs and sample days for the period between 1 January 2004 and 31 December 2011 (June, July, and August were excluded) as a function of the solar and geophysical disturbance level each day represented by Kp index, solar wind speed, solar proton flux, and solar X-ray flux. Each day is categorised into 1 of 16 possible states according to whether the threshold value for each parameter is exceeded at any time during the day. Thresholds are Kp = 2, Vsw = 450 km/s, proton flux = 1 cm−2s−1sr−1 (for >10 MeV protons), and X-ray flux = 5 × 10−7 Wm−2 (for 0.5–4.0 Å X-rays, only daytime is considered). Ones indicate that the threshold for the corresponding parameter in columns 1–4 was exceeded. Thresholds are the same as in [8]. Column 5 shows the number of sample days in each category, and column 6 shows the number of days when PMWEs were observed. Column 7 shows the percentage of sample days when PMWEs were observed. These are shown in bold type when the number of sample days is greater than 10.
1
Kp Index		2
Solar Wind Speed		3
Proton Flux		4
Short X-Ray Flux		5
Number of Sample Days		6
Number of PMWE Days		7
% of PMWE Days
00005277214
00012814
0010600
00119444
0100882528
0101800
01102150
011100-
10002718030
1001281036
10108563
10114375
110054832159
1101342162
1110252080
1111131292
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Belova, E.; Persson, S.N.; Barabash, V.; Kirkwood, S. Polar Mesospheric Winter Echoes Observed with ESRAD in Northern Sweden During 1996–2021. Atmosphere 2025, 16, 898. https://doi.org/10.3390/atmos16080898

AMA Style

Belova E, Persson SN, Barabash V, Kirkwood S. Polar Mesospheric Winter Echoes Observed with ESRAD in Northern Sweden During 1996–2021. Atmosphere. 2025; 16(8):898. https://doi.org/10.3390/atmos16080898

Chicago/Turabian Style

Belova, Evgenia, Simon Nils Persson, Victoria Barabash, and Sheila Kirkwood. 2025. "Polar Mesospheric Winter Echoes Observed with ESRAD in Northern Sweden During 1996–2021" Atmosphere 16, no. 8: 898. https://doi.org/10.3390/atmos16080898

APA Style

Belova, E., Persson, S. N., Barabash, V., & Kirkwood, S. (2025). Polar Mesospheric Winter Echoes Observed with ESRAD in Northern Sweden During 1996–2021. Atmosphere, 16(8), 898. https://doi.org/10.3390/atmos16080898

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop