2.1 Measurement platforms

Two airborne, one boat-borne, and one ground-based instrumented platforms were deployed in the vicinity of Lathuile in order to monitor humidity, temperature, wind, clouds, and aerosols in the lower troposphere over Lake Annecy and the surrounding valley environment, as well as to conduct measurements at the interface between the atmosphere and the lake, and in the lake. A brief overview of the platforms is given below, whereas the details on the instrumental payloads are given in Sect. 3:

• i.

  Airborne platforms included the following:

  • a.

    One ultra-light aircraft (ULA) was mainly dedicated to remote sensing measurements. It allowed exploring the two- or three-dimensional structure of the lower troposphere thanks to a polarized Rayleigh–Mie lidar. It also carried a Global Positioning System (GPS) device and a meteorological probe (pressure, temperature, relative humidity). This will be referred to as “aerosol ULA” (ULA-A) in the following.

  • b.

    A second ULA carried both a cavity ring-down spectrometer (CRDS) water vapour isotope analyser, a meteorological probe for pressure, air temperature, GPS and relative humidity, and a cloud water collector. The platform offers the opportunity to measure the vertical profiles of temperature, relative humidity, H²HO, H${}_{\mathrm{2}}^{\mathrm{18}}$O, and H${}_{\mathrm{2}}^{\mathrm{16}}$O, and to collect cloud water samples. The latter were collected during specific cloud flights (when meteorological conditions were favourable) and the cloud collector was opened only in clouds. We will refer to this platform as “isotope and cloud ULA” (ULA-IC) in the following.

• ii.

  For the ground-based platform, simultaneous high-resolution vertical profiles of water vapour, temperature, aerosols, and winds were acquired continuously from two co-located ground-based lidars.

• iii.

  For the boat-borne platform, an instrumented boat allowed sampling the lake water by vials at the surface film and underneath (∼ 2 m deep) for the assessment of H²HO and H${}_{\mathrm{2}}^{\mathrm{18}}$O. A CRDS water vapour isotope analyser performed measurements during 1 d at the end of the experiment just above the lake surface in parallel with the lake water sampling.

These different platforms are presented in Fig. 1 together with a view of the experiment site.

2.2 Deployment

Depending on the weather conditions, airborne platforms were deployed several times a day to document the temporal evolution of the atmospheric boundary layer over the lake. The days of operation of all platforms are summarized in Table A1 of Appendix A. The ground-based water vapour, temperature, and aerosol lidar operated continuously between 12 and 21 June in the morning (gathering over 220 h of data), while the ground-based wind lidar (WL) operated continuously between 14 and 23 June in the morning (acquiring also over 220 h of data). A total of 22 successful flights were conducted with ULA-A between 13 and 19 June, while ULA-IC performed 15 flights between 14 and 20 June. Valid cloud water samples were only obtained during the last three ULA-IC flights.

The CRDS previously installed on ULA-IC was mounted on the boat from 21 to 22 June. Boat-borne CRDS observations were made at the surface of the lake on the last outings of the boat. Nevertheless, lake water samples were taken on 14 occasions during the field campaign (see Appendix A). Finally, 28 samples of precipitation were taken during the campaign (7 of which were taken on 15 June in the afternoon) between 11 and 22 June 2019.

Figure 2Example of a typical ULA flight plan performed during L-WAIVE (performed on 17 June 2019). (a) The flight track adopted during the flight (blue dots for vertical profiling and red dots for horizontal exploration above the lake). DEM is the digital elevation model. (b) Schematic representation of the measurement strategy adopted during L-WAIVE. Two types of sampling strategy are illustrated: red for atmospheric sampling and green for cloud sampling. The purple arrows illustrate that the atmospheric sampling is performed during both ascent and descent.

Download

On days when both ULA flew in coordinated patterns (13, 16–20 June), flights typically began with a profiling sequence between the surface and ∼ 4 km a.m.s.l., which was carried out in the vicinity of the two ground-based vertically pointing lidars (see Fig. 2). Soundings with levelled legs (see dotted blue line in Fig. 2a) were performed at a relatively slow ascent rate (∼ 60 m min⁻¹) to ensure that the instruments were as close as possible to equilibrium with the environment. Upon reaching 4 km a.m.s.l., the flight route of the two ULA could differ, with ULA-A performing a high-altitude survey above Lake Annecy (see dotted red line in Fig. 2a), while ULA-IC was aiming for shallow cumulus clouds to sample cloud water droplets, as illustrated in Fig. 2b. Liquid water sampling was performed via multiple passes through the clouds to accumulate enough material to conduct isotope analysis. At the end of the flight, both ULA performed race-track descents around the ground-based lidars on their way back to the airfield.

During ascent and descent, the Airborne Lidar for Atmospheric Studies (ALiAS) aboard ULA-A was pointing sideways to directly derive the aerosol extinction coefficient (Chazette et al., 2007; Chazette and Totems, 2017). For the exploration of the valley at a cruising altitude between 3.5 and 4.5 km a.m.s.l., ALiAS was pointing to the nadir. The combination of both flight sequences thus allowed to survey the three-dimensional structure of the lower troposphere over the lake and its surroundings. The individual flight characteristics (time, maximum altitude, type of exploration) are presented in Appendix B for the two ULA (Tables B1 and B2 for ULA-A and ULA-IC, respectively).

3.1 Airborne payloads

We used two Tanarg 912 XS ULA from the company Air Création (Chazette and Totems, 2017). For each ULA (ULA-A and ULA-IC), the maximum total payload is of approximately 250 kg, including the pilot. Flight durations were between ∼ 1 and 2 h, depending on flight conditions, with a cruise speed around 85–100 km h⁻¹. The ULA location was provided by a GPS and an Attitude and Heading Reference System, which are part of the MTi-G components sold by XSens.

3.2 Ground-based instrumentation

The ground-based scientific facility hosted by the technical department of the city of Lathuile was mainly composed of the Raman lidar WALI (Weather and Aerosol Lidar) and of a scanning Doppler lidar. WALI was embedded in the ground-based MAS (mobile atmospheric station; Raut and Chazette, 2009). The UAV (unmanned aerial vehicle) was also operated from this site, close to the lidar.

3.3 Boat-borne payload: lake and atmospheric sampling

The lake water surface and subsurface were sampled at the middle of the “Petit Lac” to measure water isotopes and chemicals from the boat:

• i.

  The lake water thermal stratification was monitored using an EXO sonde, equipped with temperature, pressure, pH, dissolved dioxygen, ion conductivity, and chlorophyll sensors. Profiles recorded in the middle of the “Petit Lac” (Fig. 1) once or twice per day (Table A1) and showed that the thermocline was typically at a depth of about 7 to 13 m, in good agreement with previous studies of the lake (Danis et al., 2003).

• ii.

  Subsurface water samples were collected at a depth of 2 m in HDPE-capped flasks. Surface water samples were collected using a 30 × 30 cm silica glass plate immersed into water for a minute, then gently removed from water vertically (Cunliffe et al., 2013). The water falling from the plate in a continuous flow was not sampled; then, the dropwise water was collected in HDPE-capped flasks. The surface microlayer samples were then collected by scraping the remaining water on the glass plates (using a rubber scraper) in amber glassware capped vials. All liquid water samples were measured for isotopic composition (H${}_{\mathrm{2}}^{\mathrm{18}}$O, H²H¹⁶O) at FARLAB, University of Bergen, according to established laboratory procedures (Appendix D).

• iii.

  On 22 June, the CRDS isotope analyser was taken aboard the boat to sample a cross section in the atmospheric layer just above the “Petit Lac” through the location used for in situ sampling on the other days. This made it possible to capture the isotope concentrations as a function of the distance from the shore and the depth of the lake on that day. Water vapour isotope measurements were taken from an inlet at either ∼ 20 cm or ∼ 2 m above the water surface, while the iMet sonde measured temperature, relative humidity, and location. We used 2.5 m of $\mathrm{1}/\mathrm{4}$ in PTFE tubing and a 40 cm $\mathrm{1}/\mathrm{4}$ in stainless steel tip on the first inlet, and about 1.5 m of tubing with a 50 cm stainless steel tip. A flow rate of about 10 L min⁻¹ was provided by an external manifold pump (N022AN, KNF, Germany) to either of the selected inlet lines. Post-processing and calibration of water vapour measurements were done as for the aircraft data (Sect. 3.1.2).

3.4 Precipitation sampling

Precipitation samples were taken throughout the campaign period. The sampling device consisted in a pre-cleaned HDPE funnel directly connected to a pre-cleaned HDPE sampling bottle. The precipitation samples were manually operated: after each precipitation event, the sample was aliquoted into 1.5 mL glass vials with rubber/PTFE septa and stored at 4 ^∘C prior to isotopic analysis, while the rest of the sample was stored at −18 ^∘C. Precipitation sampling times lasted from 20 min to several hours, depending on rainfall rate.

4.1 Synoptic conditions

During L-WAIVE, France was under the influence of two main synoptic features, namely a pronounced trough over Brittany and the British Isles and a high-pressure ridge over northern Africa, sometimes extending across the Mediterranean and all the way into eastern Europe. This weather situation is illustrated in Fig. 3 using the fifth European Centre for Medium-Range Weather Forecasts Reanalysis (ERA5) at 0.75^∘ horizontal resolution. This configuration caused a particularly strong pressure gradient over the western Mediterranean for the period 12–15 June (Fig. 3a), flanked in the east by an intense Libyan anticyclone, transporting warm air from subtropics to midlatitudes. The ridge weakened and broadened over the following days (16–19 June, Fig. 3b) before strengthening again (20–23 June) as the Libyan high intensified (Fig. 3c).

Figure 3ECMWF ERA5 temperature at 700 hPa (K, shading), geopotential height at 700 hPa (gpm/1000, contours), and horizontal winds at 700 hPa (black vectors) on (a) 14 June 2019 at 12:00 UTC, (b) 17 June 2019 at 12:00 UTC, and (c) 22 June 2019 at 12:00 UTC. The location of Lathuile is indicated with a grey dot surrounded by white.

During the course of the campaign, the area of interest was under the influence of warmer temperatures in the free troposphere linked with the high-pressure ridge at the beginning and the end of the field campaign, and colder temperature from 20 to 22 June. Since the experiment was conducted in a valley, where ERA5 reanalyses are generally considered not to be very accurate below the average altitude of the mountains (∼ 2 km a.m.s.l. in our case), it is more informative to use the measurements acquired during the field campaign from the lidar and ULA to describe the evolution of meteorological variables (wind, humidity, temperature, pressure).

4.2 Insight on the local atmospheric conditions derived from lidars

During L-WAIVE, the temporal evolution of the scattering layers, including clouds and aerosols, has been monitored using both the ground-based lidar (WALI) and the airborne lidar (ALiAS). They provide insight into air masses transport, weather situation, and local convection. For instance, such a capability was used to improve our understanding of atmospheric circulation in complex situations such as extreme heat wave phenomena (Chazette et al., 2017).

Figure 4Temporal evolution of (a) the aerosol scattering ratio (ASR) where the shaded areas correspond to the presence of clouds and (b) the linear volume depolarization ratio (VDR) from 13 to 21 June 2019. The cloud type and rain location are indicated, as is the type of the main aerosol structures identified by the VDR (orange/pink for dust on 14 June and other for pollution aerosols; clouds are in pink). Ns, Ac, and Cu indicate nimbostratus, altocumulus, and cumulus clouds, respectively.

Download

Figure 4 shows the temporal evolution of the vertical profile of the aerosol (or cloud) apparent scattering ratio corrected from the molecular transmission (ASR, Fig. 4a) and of the linear volume depolarization ratio (VDR, Fig. 4b, Chazette et al., 2012). The vertical extension of aerosols in the valley is highly variable over time as is the presence and nature of clouds during the campaign. The strong diurnal cycle of ASR in the valley is tightly related to the slope winds and vertical stability associated with surface heating and the forcing due to the presence of clouds. The depth of the aerosol layer is the largest at the end of the afternoon when the convection is the strongest in the valley, while it is exhibiting a minimum around 11:00–12:00 LT when the valley winds reverse. The period was rather cloudy, with cloud types and the rainy periods including thunderstorms being indicated in Fig. 4a. In Fig. 4, we can clearly see the vertical updraft associated with clouds, which drives scatterers and water vapour towards higher altitudes, where there is a transition to the free troposphere. The transition is between the yellow and blue colours in Fig. 4a.

Figure 5Temporal evolution of (a) wind speed (m s⁻¹), (b) wind direction (0^∘ is for north and 90^∘ for east) obtained from wind lidar, and the water vapour mixing ratio (WVMR) derived from the ground-based Raman lidar. White areas correspond to missing data caused by detection limitations of the lidar (signal-to-noise ratio, presence of clouds).

Download

Significant day-to-day variations of the aerosol layer depth are thus observed during the field campaign which highlights local, regional, and synoptic-scale transport of air masses over the Annecy valley. At the beginning of the campaign (mainly on 14 June), large VDR values are detected up to altitudes exceeding 7 km a.m.s.l. (Fig. 4b). These large VDR values are related to a large-scale Saharan dust transport episode over the Lake Annecy valley favoured by the synoptic conditions on that day. These aerosols are progressively mixed downward by subsidence, reaching the valley floor around 00:00 LT on 15 June. The local wind in the valley is obviously disturbed during stormy periods and during the episode of long-range dust transport on 14 June. The presence of these aerosols at higher altitudes (up to ∼ 5 km a.m.s.l.) allowed the wind to be retrieved with the wind lidar above 2 km a.m.s.l. (Fig. 5a–b).

Strong southerly winds (in excess of 20 m s⁻¹ above 3.5 km a.m.s.l.) were observed on 14 June in agreement with the synoptic meteorological fields. Weaker winds, less than 5 m s⁻¹, are observed below 2 km a.m.s.l. for the entire period. The wind intensity does not show repetitive patterns from one day to the next. Nevertheless, it is rather in the course of the afternoon, when the wind comes from the north to the south, that we have the strongest velocities in the planetary boundary layer. Wind direction shows regular variations (Fig. 5b), with winds directed towards the south during the most of the afternoons. On the contrary, during the night and the first part of the morning, winds are directed towards the city of Annecy (north).

The vertical profiles of WVMR derived from the WALI ground-based lidar are also shown in Fig. 5c. They highlight highly variable water vapour in the first 2 km of the atmosphere, as is the case with wind. Generally higher WVMRs are often observed at night, associated with nighttime downslope winds, but may persist during days associated with heavy precipitation such as on 15 June 2019. The boundary between high and low humidity of the free troposphere is also highlighted, especially at night when the lidar exhibits enhanced sensing capabilities at higher altitude. For example, high humidity is seen up to ∼ 4 km a.m.s.l. on the night of 17–18 June in agreement with what is highlighted in Fig. 4a.

The wind at the surface of the lake controls evaporation and turbulent surface fluxes and influences the thermal convection, and cloud forcing influences the depth of the valley boundary layer, allowing the mixing of water vapour in a layer of variable thickness over time. The influence of the lake can therefore be seen up to altitudes of ∼ 4 km a.m.s.l., bearing in mind that the transition altitude to the free troposphere is generally of the order of 2.5 km a.m.s.l. here. The altitude of this sharp transition is also linked to the average altitude of the mountains surrounding the measurement site. There are no big differences between day and night on this boundary. It should be noted, however, that there are substructures that can be characterized by wind shears at different altitudes (Fig. 5b). A 0.5–1 km thick layer over the lake is thus highlighted, which can be linked to the wetter structures of Fig. 5c (in dark blue).

5.1 Atmospheric in situ sampling

In total, 15 flights with ULA-IC have been performed (see Table B2), including 13 flights where the CRDS allowed a representative sampling of δ²H and δ¹⁸O (Appendix D). The isotope content δ²H observed during the flights ranged between about −340 ‰ and −80 ‰ for flight altitudes up to ∼ 3.5 km a.m.s.l., as shown in Fig. 6a. For this figure, we considered ascending and level flight segments, and excluded samplings with pressure increases larger than 1 hPa/10 s. As observed in earlier studies, the dataset is clustered along a typical mixing line in δ²H–q space (Noone, 2012; Salmon et al., 2019; Sodemann et al., 2017). The end members of the mixing lines show substantial day-to-day variations, from ${\mathit{\delta }}^{\mathrm{2}}\mathrm{H}=-\mathrm{110}$ ‰ to −80 ‰ (∼ [−16, −12] ‰ for δ¹⁸O; not shown) for the more humid end member (q > 8 g kg⁻¹) and from ${\mathit{\delta }}^{\mathrm{2}}\mathrm{H}=-\mathrm{340}$ ‰ to −230 ‰ (∼ [−30, −20] ‰ for δ¹⁸O; not shown) for the drier end member (q < 3 g kg⁻¹). It is important to keep in mind that both the variation in maximum flight altitudes and the meteorological situation can contribute to variability.

Figure 6Overview of airborne water vapour isotope measurements. (a) Isotope content δ²H (‰) vs. specific humidity (g kg⁻¹). The mean mixing lines is computed following Noone (2012) with an end point q=1.37 g kg⁻¹ and ${\mathit{\delta }}^{\mathrm{2}}\mathrm{H}=-\mathrm{335}$ ‰, and reported using a dotted black line. (b, c) Vertical profiles of δ²H (‰). Colour indicates flight number. Panel (c) was added to separately plot the profiles with strong vertical gradients. Isotope data are averaged at 10 s intervals. The red cross associated with flight 11 indicates that no data were relevant during that flight.

Download

As shown in Fig. 6b–c, the vertical profiles of the isotope content δ²H show small variability with altitude below 2 km a.m.s.l.; the greatest variability is observed in the first few hundred metres. In some cases, the profiles exhibit strong vertical gradients at higher elevations, as seen between 2.5 and 3.6 km a.m.s.l. during five flights (flights 5, 6, 7, 9, and 10). The same patterns are observed on the vertical profiles of δ¹⁸O (not shown) and even on the WVMR profiles derived from the meteorological probes aboard the two ULA. The sharp gradients are not observed systematically, as in the case of flight 15 (20 June), or may be present at a higher altitude, not sampled by the flight.

Indeed, because of the cloud cover and depending on the wind conditions, ULA have not been able to fly in the free troposphere on most days. However, this was possible for flight 10, where a marked reduction of δ²H occurred just above 3 km a.m.s.l., while higher values of isotopic content are seen at higher altitudes that match those below the transition. The same behaviour was observed on independent measurements of water vapour mixing ratio, which may suggest filamentation associated with differential transport of drier air layer in a wind shear zone. We will discuss this vertical distribution of isotopes in more detail in Sect. 5.2.5.

5.2 Cloud liquid water sampling

Four relevant cloud water samples have been taken with the CASCC on 18, 20, and 21 June (Table A1). The isotope content of the cloud liquid water is shown in Fig. 7a (coloured circles). It is close to the global meteoric water line (GMWL) with corresponding d-excess values ranging from 12.1 ‰ to 14.8 ‰. Equilibrium condensates were estimated from relevant water vapour isotope measurements during the time when the cloud samples were taken (Fig. 7a, coloured squares) using the fractionation factors of Majoube (1971), and air temperature measurements at cloud level, ranging between −4 and 1 ^∘C. It is worth noting that given potential sources of uncertainty, the equilibrium condensate values agree remarkably well with the cloud liquid water during 18 June (green symbols) and the afternoon of 20 June (violet symbols). Overall, results here confirm that the cloud water droplets formed from equilibrium fractionation from ambient vapour. It is worth mentioning that such agreement between the completely independent sampling and measurement of vapour and cloud water supports the overall validity of (i) the cloud water sampling protocol, (ii) the airborne vapour isotope measurements, and (iii) the consistency of the calibrated water isotope dataset derived from the L-WAIVE campaign.

Figure 7δ²H−δ¹⁸O plots of liquid samples compared to the vapour isotope estimates. (a) Cloud water samples (circles), equilibrium condensate (squares), and precipitation samples (grey triangles). (b) Equilibrium vapour from lake water samples at various depths (dots and diamonds) compared to vapour isotope measurements from ULA-IC (upward triangle) and boat (downward triangle). Colour denotes matching dates (in the blue boxes are the colour legends in dot representation). Grey colours show data where the vapour samplings are not available. The times of the lake water sampling are indicated next to the dots with the same colours. The black line denotes the global meteoric water line (GMWL).

Download

5.3 Precipitation liquid water isotope composition

In total, 28 precipitation samples, of which 22 are unique, have been taken during the campaign, with sampling times lasting from 20 min to several hours, depending on rainfall rate (Table A1). The isotope composition in rainfall ranges from −11.2 ‰ to 2.2 ‰ in δ¹⁸O and −73.5 ‰ to 9.6 ‰ in δ²H (Fig. 7a, grey triangles), with corresponding values of d excess between 3.5 ‰ and 20.8 ‰ (not shown). Even if evaporation effects from the sampling setup cannot be fully excluded, in particular for the samples with a sampling duration of more than 3 h, the consistency of duplicate samples indicate that the influence of sampling artefacts can overall be considered as limited. Notably, there is an overall correspondence between the isotope range observed in cloud water samples and in a majority of precipitation samples. Deviations of precipitation samples from the GMWL indicate potential below-cloud exchange processes between the falling raindrops and the ambient water vapour, leading to an enrichment of the precipitation in ¹⁸O (Graf et al., 2019; Worden et al., 2007). We have noticed that samples taken between 12 to 15 June, and on 21 June are most enriched in ¹⁸O, exhibiting the higher values in δ¹⁸O and δ²H and a smaller d excess. These samples are from rainfall events associated with local thunderstorms. The corresponding low d excess (even negative, ∼ −8 ‰) of these samples may point to the partial evaporation of rain droplets during their fall.

5.4 Lake liquid water isotope composition

Evaporation from Lake Annecy is expected to be an important source for the water vapour in the low troposphere above the Annecy valley. In order to link the atmospheric profiles of water vapour isotopes to the lake as a moisture source, 20 lake water samples (Table A1) were taken throughout the campaign within the lake–atmosphere interface layer (6 samples), as well as close to 2 m depth (14 samples), and analysed for their stable water isotope composition. The average isotope contents were found to be −8.3 ± 1.5 ‰ for δ¹⁸O and −63.0 ± 6.0 ‰ for δ²H. To assess the isotopic balance between the lake water and the water vapour isotope composition measured by ULA and boat, we calculated the equilibrium vapour for the lake surface temperatures measured during respective days (Fig. 7b) using the equilibrium fractionation factors of Majoube (1971).

Lake water samples taken close to the surface are expected to be most affected by evaporation, causing deviations to the right from the GMWL along evaporation lines (enrichment in ¹⁸O) and corresponding to a negative d excess. As a matter of fact, lake water samples clearly cluster on the right-hand side of the GMWL (Fig. 7b, diamonds), whereas vapour data from the ULA-IC (triangles) and the boat (downward triangle) are to the left, with a relatively consistent range of isotope values. The corresponding d excess underline this result for the six lake water samples from the lake–atmosphere interface layer, with values of the d excess ranging between 2.2 ‰ and −19.6 ‰ and an average value of −6.3 ‰. For the samples taken at a depth of 2 m (Fig. 7b, diamond symbols), the derived d excess showed less evaporation influence, with a mean of 0.6 ‰ and a standard deviation of 8.2 ‰. While some sampling artefacts cannot be fully excluded, the fact that there are differences between the isotope contents at the lake surface and 2 m depth points to the impact of ongoing evaporation that decreases with depth. We note that temperature profiles taken within the lake show a typically strong thermocline below about 7 m depth during summer (Danis et al., 2003). Therefore, freshwater input from runoff and precipitation, and loss to evaporation are expected to primarily affect the warmer, upper mixed layer within the timeframe of the campaign.

The range of equilibrium vapour for δ²H observed in lake water for 16 and 17 June (Fig. 7b, blue and yellow dots) matches to first order with the range of δ²H from ULA-IC. However, the δ¹⁸O in equilibrium vapour is substantially higher than the measured vapour value. This pleads for a kinetic (non-equilibrium) fractionation during lake evaporation. During 18 June (Fig. 7b, green dot), equilibrium vapour of the liquid samples has a lower heavy isotope content than observed by ULA-IC, pointing to the influence of other factors on either atmospheric or lake water composition on these days. However, one should be rather cautious about this interpretation, as the time of the lake water sampling (∼ 09:00 LT) was significantly different from the time of the flight (flight 9 of ULA-IC, ∼ 12:00 LT), while for 16 and 17 June, the measurements were performed at the same time. It is worth noting that Lake Annecy is fed by a catchment whose surface is 10 times larger than that of the lake via 10 main tributaries located on the lake periphery. The flows of its tributaries increase significantly in situations of heavy rain, which can substantially influence the isotopic content of the lake during the course of a year; however, on 17 and 18 June, no rain was observed.

6.1 Vertical profiles in the low troposphere as observed from ULA

Figure 8 shows the vertical profiles corresponding to four coordinated flights between the two ULA. These flights were carried out in the late morning and afternoon during the two sunny days of the campaign, i.e. 17 and 18 June. The profiles were constructed by averaging the data over 100 m bins. This operation makes it possible to improve the signal-to-noise ratio and to evaluate the dispersion of the measurements as a function of altitude (coloured areas in the figure).

Figure 8Vertical profiles of the isotope composition δ²H from CRDS (ULA-IC), specific humidity (from WALI and meteorological probes aboard ULA), potential temperature (from meteorological probe of ULA-A), wind speed components from the wind lidar, and apparent scattering ratio (ASR) derived from the airborne lidar ALiAS (ULA-A) on (a) 17 June in the morning (∼ 10:30–12:00 LT), (b) 17 June in the afternoon (∼ 16:00–18:30 LT), (c) 18 June in the morning (∼ 11:00–12:30 LT), and 18 June in the afternoon (∼ 16:00–18:00 LT). The dispersion on the data is represented by the coloured areas.

Download

The vertical structures of the lower troposphere are well reproduced across all data types, showing the consistency of the isotopic profiles with the independent observations. The strong gradients on δ²H above 2.5 km a.m.s.l. are related to the transition to the free troposphere and are confirmed by the lidar observations and meteorological measurements on the two ULA. Discrepancies appear between the specific humidity measured from the two ULA on 18 June above 2.7 km a.m.s.l. (Fig. 8c). This can be explained by the presence of a fractional cloud cover. The ULA-IC has crossed part of the cloud base which is at the interface with the free troposphere.

In the morning, the atmosphere near the lake shows near-surface stability over 200 to 300 m which deteriorates/mixes away in the afternoon according to the potential temperature profiles. A few hundred metres above the lake, the atmosphere is very stable on 17 June in the morning (Fig. 8a); it then loses stability in the afternoon due to thermal convection (Fig. 8b). On the morning of 18 June, the stability is strong below 1.5 km a.m.s.l. (Fig. 8c) and approaches the neutrality above this altitude. In contrast to 17 June, an atmosphere close to neutrality is observed in the afternoon on 18 June (Fig. 8d), which may favour vertical exchanges from the lake surface to the free troposphere. An important difference between the 2 d is also related to the surface wind speed. Winds of the order of 5 m s⁻¹ were present on 17 June, whereas they were close to null on 18 June. Therefore, significant differences in water exchange between the lake and the atmosphere can be expected between the 2 d. These differences have already been noted when comparing the samples in Fig. 7b. On 17 June, wind-forced evaporation phenomena may therefore intervene which may explain at least part of the assumed isotopic fractionation. On the other hand, on 18 June, we have seen that the water samples are hardly representative for ULA measurements as they were taken early in the morning, before the sun lit up the valley. Around noon, the structure of about 300 m observed at the bottom of the δ²H profile (Fig. 8d) and found in the other data is certainly strongly influenced by evaporation from the lake. It is significantly attenuated in the afternoon due to vertical mixing linked to thermal convection. We observe an enrichment in ²H in this layer compared to the rest of the vertical profile in the morning that may be due to the lake evaporation into depleted air.

6.2 Statistical overview of isotopic content from the lake to the atmosphere

The main atmospheric vertical structures are derived from the profiles of potential temperature, aerosols, wind, and water vapour mixing ratio. We have therefore defined three layers in the low troposphere: (i) between the lake surface to 1 km a.m.s.l., (ii) between 1 and 2.5 km a.m.s.l., and (iii) between 2.5 and 3.5 km a.m.s.l. The uppermost value represents our altitude sampling limit. The first layer can be identified as the lake boundary layer, the second layer is associated with the region of influence of the lake, below the altitude of the surrounding mountains, and the third layer is the transition towards the free troposphere influenced by the regional circulation. The location in altitude of these different layers obviously depends on the weather conditions and the time of day.

Figure 9Whisker boxes for (a) d excess, (b) δ²H, and (c) δ¹⁸O. For each subfigure, the upper part depicted the statistical content of isotope in water vapour and the bottom part in the liquid water. For each whisker box, the central continuous vertical line indicates the median (50th percentile, P₅₀), and the bottom and top edges of the box indicate the 25th (P₂₅) and 75th (P₇₅) percentiles, respectively. The whiskers extend to the most extreme data points not considered outliers, which are values outside the interval bounded by the 5th and 95th percentiles. For the atmospheric water vapour, the number of samples n is indicated close to the boxes. The confidence interval at 95 % defined by $\left[\right(P_{\mathrm{50}}-\mathrm{1.57}\cdot (P_{\mathrm{75}}-P_{\mathrm{25}})/\sqrt{n}),(P_{\mathrm{50}}+\mathrm{1.57}\cdot (P_{\mathrm{75}}-P_{\mathrm{25}})/\sqrt{n}\left)\right]$ is represented by the notches in the upper part of the subfigure.

Download

Using the vertical layering introduced above, the overall results for stable isotopes in water are summarized in Fig. 9 on a statistical basis selecting the highest-quality data. This synthesis is presented in the form of whisker boxes for d excess (Fig. 9a), δ²H (Fig. 9b), and δ¹⁸O (Fig. 9c). For each subfigure, the upper part depicted the statistical content of isotope in water vapour for each sampled altitude range in the atmosphere, while the lower part describes the same thing for liquid water samples collected from clouds, precipitation, and in the lake, at the interface and subsurface. For the atmospheric water vapour, the sample number n is significant (numbers given at the right of the whisker boxes in Fig. 9b), and it makes sense to accurately assess the confidence intervals at 95 %. As the notches do not overlap from an atmospheric layer to another, we can conclude, with 95 % confidence, that the true medians do differ. Hence, we highlight a significant variability in the stable isotope content of water vapour depending on the atmospheric layers previously identified.

Between the surface of the lake and 2.5 km a.m.s.l., we find δ¹⁸O values are in the lower range to those recorded by Craig and Gordon (1965) for evaporation over the Mediterranean [−15, −10] ‰. On the other hand, they report much more dispersed d-excess values ranging from 5 ‰ to 35 ‰. Gat (2000) gives narrower values for marine and European air masses with a d-excess interval of [7–11] ‰. Our observations are mainly outside this interval for the first two layers; they overlap it for the 2.5–3.5 km a.m.s.l. layer more influenced by the synoptic scale. Measurements from the boat on the last day of the campaign within 1 m above the lake are overall consistent with the vapour isotope measurements made by ULA-IC on previous days in the lake boundary layer.

Based only on the statistical information content provided by the whisker boxes profiles, it is difficult to see precisely whether the evaporation of lake water significantly influences the first layers of atmosphere just above the lake. Indeed, different factors can influence the relation between water vapour measured by the ULA above the lake and the evaporating water vapour. In addition to the immediate evaporate, water vapour can be advected and mixed in from other locations horizontally but also vertically. In particular for a lake with coastal, vegetated areas within less than about 2 km distance anywhere on the lake, influences from evapotranspiration are likely within the lower 1 km above the lake.

Our liquid water samples are not numerous enough to define a confidence interval. As explained previously, we nevertheless note an enrichment in heavy isotope of the surface layer of the lake which is directly in contact with the atmosphere and therefore directly subject to evaporation. The 2 m depth layer is significantly less enriched as evidenced by their isotope content. We highlight therefore a significant enrichment in heavy isotopes of surface water associated with low values of d excess, which is characteristic of evaporation as described by Gibson and Edwards (2002). Measurements of δ¹⁸O profiles in Lake Annecy have already been carried out between the years 2000 and 2002. They have shown δ¹⁸O values of the order of −9.5 ‰ for water above the thermocline of the “Petit Lac” during the summer months (Danis, 2009), in agreement with our results. This indicates a year-to-year stability of the isotopic content of the lake water below the surface microlayer.

From a first analysis of our successful airborne in situ sampling of water vapour isotopes and cloud water, we conclude that cloud water formed in an equilibrium fractionation process for the samples collected here, and the cloud water isotopes were relatively similar to local precipitation (taken on other days). Since 1992, the Global Network of Isotopes in Precipitation (GNIP, https://nucleus.iaea.org/wiser, last access: 5 July 2021) regularly samples rainwater at Thonon-les-Bains (60 km northeast from Annecy) for isotope analyses. Based on data available until 2018, typical summer δ¹⁸O (δ²H) values fall within the interval [−16, −1] ‰ ([−120, −20] ‰), encompassing our precipitation and cloud water measurements. It should be noted that the lake water is poorer in heavy isotopes than rain and cloud water. This difference is probably due to snowmelt during spring and the contribution of precipitation falling at high altitude. Note that GNIP data show median values of −10.5 ‰ (−78.6 ‰) for δ¹⁸O (δ²H) during winter versus −6.5 ‰ (−43.5 ‰) during summer which agree with the studies of Dutton et al. (2005) and Kendall and Coplen (2001).