Month 36 Interim report 4

WP2 (The Carpathians)

1. Remote sensing and in-situ topography in alpine and landslides sites

Phase 2023

In-situ ground-based DGPS measurements were performed in the summer and autumn field campaigns in order to set an accurate location of the borehole and ground control points for the LiDAR and UAV surveys. Two filed campaigns, in February and April 2023, have been conducted in order to asses snow accumulation on the Galesu rock glacier, in Retezat Mountains.  The aim of the snow measurement is to better understand the winter conditions that favour the cooling of the ground and thus the preservation of permafrost. Also, the CryoGrid permafrost evolution model is an important part of a plan to apply for future projects with the Oslo partners, and the multi annual snow proprieties (e.g. thickness, density) are an important input factor for it.

Using UAV photogrammetry and surface from motion (SfM) method two digital surface models (DSM) were constructed, and each one was compared against a DSM of the same area in snow free conditions. The drone’s (DJI Phantom4) flight path has been pre-planned with terrain characteristics, camera parameters and mission duration required for the targeted spatial resolution of images, using DJI Ground Station Pro and Dronelink software with an Elite licence. The resulting DSM have been calibrated by ground control points (GCP) acquired using a differential global positioning system measurements (DGPS) setup comprising of a base and rover stations with real time kinematics (RTK), in order to ensure centimetre accuracy.

Figure 1 – A comparison between the Orthomosaic, DSM (digital surface models) and terrain shading with snow cover (winter and early spring) and without snow cover (summer) on Galeșu rock glacier (Retezat Mts). The snow accumulation can be observed in all three types of images

To further validate the results, snow thickness measurements have been performed using graded avalanche probes. Following the field data acquisition, the process of generating the DSMs was carried out using Agisoft Metashape Professional software (Fig. 1). Due to the relatively difficult process of finding commune reference points in the images an overlap of at least 80% has been used.

Digital Elevation Models generated from second-year drone-based photogrammetry and terrestrial laser scanner (TLS) of Siriu dam debris-flow, Bâsca landslide and Chirlești earthflow were used to derive additional displacement rates by comparison with 2015-2020 data from PSInSAR analysis performed by Terrassigna Inc. (https://pstool.terrasigna.com) using Sentinel-1 SAR. The purpose of this analysis is to provide a more comprehensive estimate on surface velocity, to be integrated in run-out models which are being developed for these study sites. High-accuracy Digital Elevation Models were generated from multiple LiDAR flights over the complex deep-seated landslide – river system in Păltineni (Curvature Carpathians) and were integrated in the structural connectivity analysis (Fig. 2). The DEMs were used to distinguish between the different geomorphological sectors of the landslide which would have been otherwise difficult to identify due to the dense forest cover. Point cloud classification was carried out to distinguish ground points from forest canopy using the CSF plugin in CloudCompare software.

Figure 2 – UAV LiDAR hillshade DTM of Păltineni deep-seated landslide and morphodynamic sectors: accumulation/depletion A—C); different magnitude scarps (recent—D; old—E; active—F); landslide deposit reactivations (river-induced—G; non-river-induced—H); erosional forms (gully—I, steep bare slope, with sheet wash associated with rock falls—J); and fluvial forms (K, L) (from Micu et al., 2023)

2. Geochronology of landforms and dynamics reconstruction

Phase 2023

Absolute dating of landform surfaces. In November 2023 a field campaign was organised with the aim of collecting rock samples from five rock glaciers in Valea Rea, Retezat Mountains. The rock samples are to be processed by P3 in order to assess the formation and evolution of the periglacial features in the valley. Depending on the rock structure, two or three samples were collected from 11 locations. The samples are well distributed to cover the highest part of the valley, where there is still permafrost, and also to form a vertical transect of the valley from around 2200 m a.s.l. to around 1600 m a.s.l. The vertical transect assures that all the forming stages of the rock glaciers will be dated. This comes to supplement the existent database with one of the most interesting and well-developed rock-glacier in the Carpathians.

Additional rock samples were also collected from Lăițel and Capra Valleys from two rock glaciers/moraine landforms in order to enlarge the database needed for the deglaciation chronology reconstruction in Făgăraș Mountains. In Lăițel we focused on a multiple generation landform where we sampled 5 boulders. Three additional samples were collected from the upper part of Capra rock glacier with multiple parallel crests (Fig. 3).

 

Figure 3 – Rock sample collection sites in Lăițel and Capra landforms from Făgăraș Mountains

Sample processing for absolute age dating. In this phase of the project, Cosmo lab RoAMS at P3 (IFIN-HH) assured the physical and chemical processing of 105 samples, as follows: 68 samples from Făgăraș Massif, 14 samples from Buzău Mountains area (landslides) and 23 samples from Retezat Mountains. For documenting Younger Dryas and Early Holocene glaciers formation in the upper valleys of Făgăraș Massif (Southern Carpathians, Romania), a total of 47 new numerical ages (¹⁰Be exposure dating) were measured. All the sampled sites are restricted to cirque surfaces (Fig. 4). The spatial distribution of exposure ages covers the central high sectors with a prevalence of the northern and eastern exposures due to their favourability for glacier development at the end of the Late Glacial.

Figure 4 – Examples of sampling sites in Făgăraș mountains: A – Urlea Valley, B – Hanging Urlea Valley, D – Mioarele Valley

Assessing the post-LIA response of geomorphic hazards by dendro-geomorphology

Dendrogeomorphology analyses were continued in order to reconstruct past occurrences of snow avalanches and rockfalls. Snow avalanches were studied in the Făgăraș Mountains and rockfalls were subject to analysis in the Retezat Mountains in 2023. Past process history was resumed in century-long chronologies and process frequency was calculated for different time scales. In both cases, we have noted an increase in the frequency of high-magnitude events over the past decades, indicating that climate conditions are increasingly conducive to their incidence in the Southern Carpathians.

We have reconstructed the snow avalanche regime of the last century in the central part of the Făgăraș Mountains by applying a multi-path tree-ring based approach. Results highlight a number of 16 winters with widespread show avalanche activity across the region (Fig. 5).

Figure 5 – R-It based snow avalanche year reconstruction for the study area (Făgăraș chronology). Event years with R-It values exceeding the 10% threshold are marked with red; event years with R-It values between 7% and 10% are marked with yellow

4. Climatic data analysis

Phase 2023

Climate downscaling over the Southern Carpathians. After completing the development of the climate downscaling toolbox TopoPyScale in 2022, we were able to apply it over the SC range for the 1950-2020 ERA5 dataset archive. All variables required to drive permafrost or snow models were downscaled at a ground resolution of 100 m. This includes the hourly variables air temperature, precipitation, air pressure, air humidity, incoming shortwave and longwave radiations as well as wind.

Figure 6 – Mean air temperature for the period 1960-1980 (top panel), 2000-2020 (mid panel), and the difference between both. Temperature increased more in the lower elevation and mid-elevation (+1.5-2^oC) than higher elevation (+1^oC). We can also observe the reduction of below freezing temperature in blue (top panel)

The primary work was to validate the downscaled product by 1) comparing air temperatures to local weather station observation (i.e. Bâlea Lac, Țarcu, Vârfu Omu), 2) comparing temperature trends to previous studies in the regions, 3) simulating snow accumulation with the Factorial Snow Model (FSM, Essery et al. 2015) and comparing to local observation and 4) comparing downscaled variable to the available and overlapping gridded dataset (i.e. CarpaClim, ROCADA, and E-Obs) for the period 2000-2010. Figure 6 shows an example of air temperature changes of the SC range.

Continuous in-situ monitoring of climatic parameters and permafrost thermal state. The automatic meteorological station installed in September 2022 failed to provide data due to malfunction caused by potential vandalism. P2 and P1 team members re-visited the site in September 2023 in order to resume the station functionality. The station was repaired and maintained for a new winter season of measurements. First data will be collected in summer 2024 if all goes as planned. The station records air and ground temperatures, air pressure and humidity, wind speed and direction, incoming shortwave radiations, and snowdepth.

In 2023 we continued to monitor ground surface temperatures (GST) at 19 sites in the Retezat Mountains using miniature dataloggers. At most of the monitoring sites, the mean annual ground surface temperature (MAGST) in the hydrologic year 2022/2023 revealed a positive trend. At PIE and VREA sites, for example, MAGST in the last monitoring year was with more than 0,5˚ C higher than the mean of the interval 2012-2023 (Fig. 8). Excepting GAL site, both the winter and summer temperatures were considerably higher than the average value for the last decade (Fig. 9).

Figure 7 – View of the weather station location after refurbishment, ready for a winter season in September 2023

Figure 8 – MAGST evolution between 2012 and 2023 at four sites in the Retezat Mountains

Figure 9 – Mean daily ground surface temperature at four sites in the Retezat Mountains between 2012 and 2023

In December 2023 a field campaign has been organized to drill a 6-meter-deep borehole in the upper part of the Doamnei Valley, in Făgăraș Mountains. The location has been chosen because of 1) its relative accessibility, being in the vicinity of the Transfăgărășan road; 2) its location in an area with permafrost; 3) the existence of previous measurement that include geophysical soundings, terrestrial laser scanning and multiannual thermal measurements.

The drilling has been executed using a Shaw core drill. The borehole has been equipped with 7 S-TMB-M017 Hobo temperature sensors (1m apart), connected to two external H21 data logger, that will record the ground temperature at different depths.

The data loggers are set to record the temperature every hour. Another data logger has been set to register the ground surface temperature in the vicinity of the drill site. The climatic data (temperature, precipitation, snow depth, short wave radiation) recorded by the weather station set up in the same upper valley in September 2022 will help for the interpretation of the temperature data in the borehole.

Figure 10 – The location of the Weather Station installed under the ClimaLand project (1), the borehole site drilled under the ClimaLand project (2) and their position relative to the Balea Lac Complex and Balea Lac weather station (3). In the medallion, lower right corner of the image, the position of V. Doamnei in the Fagaras Mountains is marked with a red square

5. Complex modelling for hazard assessment

Within this task of the project, the BTS-derived model of permafrost distribution in Southern Carpathians was finalised and a joint manuscript depicting it was sent and accepted for publication in Permafrost and Periglacial Processes journal. The paper offers a comparison between the eight machine learning algorithms involved in the creation of the final ensemble model based on their median values. It discusses the environmental factors involved in sporadic permafrost distribution where talus deposits availability quantified by NDVI seems to be the primary factor of favourability along with altitude and slope aspect. A total permafrost area of about 19 km² is estimated to occur in Southern Carpathians from which most is found in the granitic massifs of Retezat and Parâng. The highest permafrost probability is recorded in rock glaciers and talus slopes from shaded and high-altitude glacial cirques. A conceptual model of permafrost distribution in Southern Carpathians is represented in Figure 11.

Figure 11 – Topographical profile on central area of Retezat Mountains indicating several debris landforms and the potential permafrost distribution pattern according to the BTS measurements.

WP3 (Danube delta and Lower Danube floodplain)

1. Cores and drills in Lower Danube Valley and Danube delta and their response to sea-level rise during the Holocene
2. Time-series analysis and reconstruction of the Lower Danube Valley and DD response to sea-level rise during the Holocene

Here we based on a network of 19 cores and four deep drills covering the Lower Danube Valley and Danube delta (Fig. 12). The coring was performed by Cobra percussion corer and usually reached the depths of 10-17 m while the drills were made by a professional company (Tehnofor Star SRL) using a drilling machine which sampled as deep as 35 m. The use of the deep drills followed our purpose for recovering a full stratigraphy covering the Last Glacial Maximum – Late Holocene interval. All the boreholes and drill used the same methodology for getting robust chronologies (using 14C ages performed at our partner IFIN-HH) and sedimentological analyses (grain-size parameters: mean-size and sorting, organic matter and inorganic carbonate content by LOI) whilst meiofauna (ostracoda) composition was identified (determined) for five of them (Pardina, Somova, Crapina, Brateș, Siret) to track the spatial coverage of the Maximum Flooding Surface (MFS). Supplementary, macrocharcoal analysis were performed on two cores (Somova and Jijila) in order to assess the human presence during different (pre)historic time periods.

Figure 12 – New cores and drills made in the Lower Danube Valley and Danube delta.

3. Estimation of the delta plain changes

Our research in the frame of this purpose (Estimation of the delta plain changes) focused on assessing sediment accumulation rates in lakes and levees in the Danube Delta, employing geochemical analysis and measurements of radionuclides 137Cs and 210Pb. Thus, 10 short cores (1-1.5 m) have been retrieved from several lakes but also from different river branches and secondary channels in the Danube delta using a van Beeker sampler (Eijkelkamp) which is a tool dedicated to undisturbed extraction of lacustrine/aquatics sediments (Fig. 13).

Figure 13 – Sampling sites of sediment cores for deltaic sedimentation rates assessment

WP4 (Comparative assessment of landforms dynamics and geo-hazards sensitivity to climate change and human pressure in the Carpathian-Danube-Black Sea system as ground for the climate-humans-landscape changes analysis)

1. Coupling climate conditions with landforms dynamics and sediment fluxes at different time scales

Reconstruction of glacial advances and deglaciation in Făgăraș mountains. Based on the results from WP2, during this phase of the project we developed a landscape evolution scenario in the formerly glaciated areas of Făgăraș mountains, with the aim to identify the response of cryospheric elements to past climate changes which is presented as follows, chronologically.

According to the obtained LGM results from Făgăraș Massif (Doamnei Valley – 19.1±2.3 kyrs and Capra Valley – 19.4±2.2 kyrs), we exclude the existence of a local Last Glacial Maximum, as suggested in the Rodna Massif by Delia Gheorghiu in her doctoral thesis presented in 2012 (37 kyrs). In terms of altitude, there are differences between the two valleys. Although the exposure ages to cosmic radiation are similar, in the southern sector, the sample collected from Capra Valley is located at an altitude of 1220 metres, while in the northern sector, the sample taken from Doamnei Valley is situated at 1480 metres. These differences arise from the varying slopes of the two macroforms. The Oldest Dryas stadial is characterised by two phases: one between 18 – 16.1 kyrs and the other between 16 – 15 kyrs. In the first phase of the stadial, the results are from the glacial threshold located in Doamnei Valley, at altitude of 1590 metres, has an exposure age of 17.7±2.3 kyrs, while the other two samples with similar exposure ages (16.8 kyrs) are located in Podrăgel Valley and Mușeteica Valley. In Podrăgel Valley, the sample is taken from the moraine that dams Lake Podrăgel, situated at an altitude of 2069 metres, and presents an exposure age of 16.8±0.47 kyrs. The sample collected from Valea Mușeteica is located at an altitude of 2136 metres (Fig. 14).

Figure 14 – Reconstructed LGM glaciers for southern (left side) and northern slope of Făgăraș Mountains (right side)

The second glacial advance during the Oldest Dryas stadial occurred at the end of the stadial. According to our results, a clustering of exposure ages is observed, ranging between 15 – 16 kyrs. In terms of altitude, these samples are grouped between 1850 and 2129 metres. This glacial advance is encountered in the following glacial valleys: Fundul Caprei, Paltinu, Dejani, Langa, Urlea, Arpașul Mare, Văiuga, and Bâlea. This glacial advance synchronises with the generation of moraines at altitudes of 1850 – 1950 metres in the Retezat Massif, which presents similar exposure ages: 15.2 ± 0.7 kyrs (Ruszkiczay – Rudiger et al., 2016). During the Older Dryas stadial, there were indications of minor glacial advances within the cirques of the Făgăraș Massif. The subsequent Allerød interstadial was characterised by a period of warming, although not as pronounced as the Bölling interstadial. This allowed for the persistence of glaciers formed during the Older Dryas in the shaded cirques at higher altitudes. In this study, we identified the existence of glaciers formed during the Older Dryas and these persisted until the end of the Allerod interstadial in glacial valleys such as Dejani (with 3 samples averaging an exposure age of 13.3±0.35 kyrs), Sâmbăta (13.3±0.37 kyrs), Podrăgel (13.1±0.36 kyrs), Arpașul Mare (13.6±0.39 kyrs), Văiuga (13.2±0.38 kyrs), Bâlea (13.6±0.40 kyrs), and Paltinu – Căldarea Berbecilor (13.4±0.38 kyrs).

Figure 15 – Glacial chronology in Făgăraș Mountains

The results obtained in this study certify the existence of Younger Dryas glaciers within the Făgăraș Massif. Out of a total of 134 processed samples, 16 samples fall within the Younger Dryas range, 5 slightly exceed the upper limit of the Younger Dryas, up to 13.3 kyrs, while 11 samples fall within the range of the Early Holocene (Fig. 15).

The abundance of Early Holocene records (18 samples) raised the questions about rock slope failures events that followed the deglaciation by intense rock wall permafrost degradation (Vasile et al., 2022), but the short period called Preboreal Oscillation (Bjorck et al., 1997) was cold and humid and could have generated small advances and formed glacierettes, as a response to the climatic forces over Europe. We can argue that in Early Holocene former glaciers from Făgăraș Mts, have been affected by cold climatic conditions initiated by PBO and continued until 9.3 kyrs event.

2. Human impact on landforms dynamics

Analysing the impact of infrastructure and hydro-technical constructions on landforms dynamics in Buzău Mountains. During this phase of the project, within this task we focused mainly on the effect of human interventions management as potential landslides triggering factor, by integrating data such as landslides inventories and susceptibility analyses. We investigated the evolution of Bâsca Rozilei landslide, which occurred in June 2021, as an example of deficient infrastructure management, which could be prevented in similar contexts, given proper prevention measures. In this purpose, we made a preliminary analysis of the landslide morphogenesis, predisposition, preparing and triggering factors, based on UAV surveys (Fig. 16) and short-term monitoring, and we also estimate a set of measures meant to reduce potential short-term damage in this particular or similar cases.

Figure 16 – Aerial view of Bâsca Rozilei landslide

The landslide occurred at the contact between Paleogene sandstones and marly-schists, at the confluence between Buzău and Bâsca Rivers, in Vrancea seismic region, Buzău Mountains, and it affects the upper and central sectors of a south-exposed slope over an old landslide deposit, partially reactivating it. This was a deep movement which mobilized 10-20 m thick rock strata. The triggering factor is represented by precipitations (cumulated quantities of over 16 mm in the previous 5 days before the slide and 45 mm in the previous two weeks). Nevertheless, the lack of long periods high precipitations (over 10 mm in 24 hours) with almost two weeks before landslide occurrence, and the lack of similar processes reported in the area led to the hypothesis that the trigger of the event was less important than its preparing causes. Further investigations revealed that for a long time period, in the upstream of the parallel river there was a directional discharge of water evacuated from the gallery under construction of Surducu water transport system towards Siriu hydropower dam. The almost continuous supply of a supplementary water volume led, in time, to the overload of the latent landslide body, contributing to its reactivation as a deep landslide.

In such contexts draining the upper surface and even the deep-water flows outside the landslide system should be performed in such manner that the redirecting of an additional water supply along a new draining axis does not cause another disequilibrium within the slope stability, water channel or nearby infrastructures.

3. Adaptation measures to future challenges in the Carpathian-Danube Delta-Black Sea area in the context of global warming and increasing human pressure

The Danube Delta-Black Sea area is particularly vulnerable to the impacts of global warming, including increased storminess, sea level rise, and coastal erosion and inundation. These changes threaten ecosystems, human settlements, and economic activities in the region. Consequent to the analyses performed during the previous phases of the project, we propose 10 adaptation measures that should be implemented to mitigate these impacts:

1. Wetland Restoration: Restore and expand wetlands in the Danube Delta area. These ecosystems act as natural barriers against storm surges and flooding, absorbing water and reducing the energy of waves.
2. Sustainable Water and Sediment Circulation and Management in the Danube Delta: Implement monitoring infrastructure and modelling frameworks for water and sediment circulation inside the Danube Delta and on its open coast in order to sustain informed actions and policies.
3. Sustainable Water Management: Implement integrated water resource management strategies that account for increased variability in water availability and demand, including the use of rainwater harvesting and the construction of water storage facilities.
4. Controlled Flooding Areas: Designate areas for controlled flooding to absorb floodwaters during extreme events, reducing pressure on levees and other flood defences. These areas can also provide valuable habitats for wildlife.
5. Adaptive Land Use Planning: Periodically update local and regional zoning regulations and development guidelines to discourage construction in high-risk areas, encourage the use of flood-resistant building techniques, and protect natural buffers like wetlands and forests.
6. Monitoring and Early Warning Systems: Implement advanced monitoring systems for weather and sea conditions to provide early warnings for storms, flooding, erosion, rip currents, algal blooms, pollution etc. Ensure that these systems are accessible to all communities, including remote areas.
7. Community Awareness and Preparedness Programs: Increase awareness about the risks of climate change and promote community-based adaptation strategies. Develop emergency response plans and conduct regular awareness campaigns for extreme weather events.
8. Sustainable Tourism and Economic Diversification: Develop sustainable tourism practices that minimize environmental impact and promote economic diversification to reduce the region’s vulnerability to climate-related changes in natural resources and ecosystems.
9. Enhanced Coastal Defences: Strengthen and elevate dikes, sea walls, and revetments (where already present near critical social and economic infrastructure) to protect against storm surges and sea level rise. Incorporate as much as possible natural solutions like ”building with nature” or ”living shorelines”.
10. Climate-Resilient Infrastructure: Design and construct infrastructure (roads, bridges, buildings) with elevated foundations and materials that are resistant to flooding and storm damage. Prioritize critical infrastructure for early adaptation efforts.

Implementing these adaptation measures requires a coordinated effort among government agencies, local communities, national and international partners. It also demands significant investment in research, technology, and infrastructure development. Adaptive management strategies should be flexible and responsive to evolving climate scenarios and scientific understandings, all these being communicated to local authorities and stakeholders with the occasion of different dissemination events.

Month 28 Interim report 3

WP2 (The Carpathians)

Remote sensing and in-situ topography in alpine and landslides sites

Phase 2022

In-situ ground-based DGPS measurements were continued in 2022 in Retezat Mts (Galeșu valley) and Curvature Carpathians using Topcon Hiper V and Leica Viva high accuracy DGPS systems for establishing reference ground points in UAV-surveyed sites (PP, P2 postdocs). High-accuracy Digital Elevation Models were generated from second-year drone-based photogrammetry and terrestrial laser scanner (TLS) of Doamnei rockwalls-RG systems, Siriu dam debris-flow and Bâsca landslide. Repeated TLS measurements were performed during the common field campaign in September-October 2022 in Doamnei glacial cirque and Siriu dam debris flow locations (PP, P1, P2). This allowed the first point clouds comparison and displacement rates analysis. UAV-surveys were also repeated for Siriu dam and Bâsca Rozilei landslides for multi-annual comparison and runout modelling. Additional data were derived in October 2022 from a LiDAR flight over the complex deep-seated landslide – river system in Păltineni (Curvature Carpathians) and from UAV in Chirlești earthflow (PP, P1), with potential effect on National Road 10, for runout modelling and connectivity analysis.

In marginal periglacial environments, such as the Southern Carpathians, where the kinematics of the rock glaciers are defined by slow flow, specific remote sensing techniques (e.g., SAR-based persistent scatterers interferometry) are preferred because they are capable of providing ground displacement accuracies on the order of a few mm. In this phase of the project, PSInSAR analysis performed by Terrassigna Inc. (https://pstool.terrasigna.com) using Sentinel-1 SAR images acquired between November 2014 and October 2020 was used to assess the kinematics of rock glaciers in the Retezat Mountains. Based on the multi-annual surface velocity rates moving areas (MA) were identified and manually delineated within the inventoried rock glaciers.

Geochronology of landforms and dynamics reconstruction

Relative dating of landform surfaces. Schmidt hammer measurements for relative dating of glacial and periglacial landforms (PP) performed over the three summer-autumn seasons (2020-2022) were processed and statistically analysed. R-values were measured on 8 RG in Retezat (Fig. 1) and 5 RG in Făgăraș, where samples were also collected for absolute dating, on the terminal fronts and interior crests of complex RG. Mean values were then correlated with both location of the investigated landforms and with the already available absolute ages (Fig. 2) and presented during the last project workshop in Norway (August 2022).

Figure 1 – Schmidt Hammer and TCN dating sampling sites in the central area of Retezat Mountains

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 2 – Altitudinal distribution of average rebound values (R) on North-exposed rock glaciers from Retezat Mts (up) and correlation with surface exposure ages obtained in Făgăraș and Retezat (bottom)

Absolute dating of landform surfaces. During 2020-21, a total of approximately 130 samples were collected for absolute age determination by Terrestrial Cosmogenic Nuclides (¹⁰Be, ³⁶Cl) surface exposure dating from glacial and periglacial sites covering 14 glacial valleys/cirques (76 samples) and 12 RG (40 samples) in Retezat and Făgăraș Mts. Depletion areas of 4 deep-seated landslides in the Curvature Carpathians were sampled, including inheritance reference sites for each (14 samples). The samples were processed for ¹⁰Be and ²⁶Al dating by P3. Final results were delivered and have completed the database for glacial landforms and periglacial landforms from Făgăraș Mts. (June 2022). Most samples from periglacial landforms in Retezat Mts. and landslides (Curvature Carpathians) collected in 2021 were lab-processed during 2022 and are currently in advanced stage for final measurements. Additional samples from Făgăraș Mts were also collected in October 2022 (Capra rock glacier).

Analysis of inter­nal structure of debris deposits by drilling and coring supplemented by geophysical sounding

In April 2022, a new sediment core of 9.5 m was obtained from Bâlea Lake (central Făgăraș Mts., 2034 m a.s.l.) using a percussion drilling set (PP) set on the frozen lake surface. Sedimentological and grain size analyses, magnetic susceptibility, radiocarbon dating, loss on ignition and C/N ratios determinations are presently being applied on the core (CO and P3 labs). Data will be used to derive age-depth models to characterize the past activity of debris flow and to indirectly infer the Holocene climate variability in the area.

Assessing the post-LIA response of geomorphic hazards by dendro-geomorphology

The dendrogeomorphological studies continued in 2022, with a focus on rockfall and snow avalanches. Building upon the previous year’s research findings in the Făgăraș Mountains (published in 2022, see Fig. 3), we investigated a rockfall site in the Retezat Mountains. In August 2022, we extracted 185 increment cores from 66 trees affected by rockfall, and processed them to establish a century-long event chronology. Given the distinct geological features, we plan to conduct a comparative analysis between the sites in the Făgăraș and Retezat Mountains to examine the different process behaviour induced by the geological setting.

Figure 3 Rockfall activity reconstruction in the Făgăraș Mountains (Sâmbăta valley)

Regarding snow avalanches, we employed a multi-path tree-ring approach in the Făgăraș Mountains. By examining 17 distinct paths, we were able to construct a high-resolution local-scale chronology of snow avalanche occurrence spanning the last century.

Climatic data analysis

Phase 2022

The downscaling toolbox in Python developed for mountainous regions using global reanalysis models such as ERA5 ranging from 1950 to present has been finished in 2022 and successfully tested with data from mid and high-altitude meteorological stations in the Carpathians. The algorithm is currently used for trend analysis and generation of climatic forcing dataset for permafrost and snow models within the project.

In September 2022 an automatic meteorological station was installed at 2100 m a.s.l. on the western flank of Doamnei glacial cirque for long-term monitoring of microclimate in this reference site for permafrost monitoring in the Southern Carpathians (Photo 1, PP, P2 with assistance from P1). The station uses a CR1000X datalogger (Campbell Scientific), a sonic anemometer, air temperature, relative humidity, pressure, shortwave solar radiation, liquid precipitation, snow depth and three soil temperature sensors which have been setup to measure every 10 minutes.

Both in 2021-2022 field campaigns the existing network of ground temperature measuring sensors within rock glaciers has been maintained for assuring continuing measurements (P2, PP), and additional itinerant BTS measurements were made to improve the coverage of permafrost distribution models (P2). In addition, the BTS measurements in the Retezat (Fig. 4) and Parâng Mountains continued in March 2022, while snow volume measurements by UAV and snow depth probing were initiated in early February 2023 (P2).

Figure 4 – Results of BTS measurements on four different rock glaciers in the Retezat Mts in March 2022

Photo 1 – The automatic meteorological station set in Doamnei glacial cirque (Sept 2022)

Complex modelling for hazard assessment

For landslide susceptibility models for hazard scenario development at regional scale we have built the computer code and mathematical relations for landslide susceptibility models and numeric simulations and collected in situ data for validation. For a representative (morpho-litho-structural and landslides typology points of view) area situated at the contact of the flysch mountains with molasse hills (Pătârlagele, Nehoiu and Siriu municipalities), susceptibility maps were produced using two methods: machine learning (random forest) and multivariate statistics (multiple logistic regression). A total of 11 predictors were used. Eight predictors (elevation, slope, texture, general curvature, profile curvature, plan curvature, topographic position index, and roughness index) were extracted from a DEM (digital elevation model) with a spatial resolution of 30m. Another three predictors (NDVI – normalize difference vegetation index, NDWI – normalize difference water index, NDMI – normalize difference moisture index) were extracted from Landsat8 OLI satellite images with the same spatial resolution. The models were trained on a data set based on an inventory of landslides. The used data set consisted of 400 points located on landslide scarps and bodies and 400 points located on terrain that was not affected by landslides.

Digital elevation models were obtained for four study sites and in-situ temperature and precipitation monitoring started. Runout modelling for Siriu dam debris flow and Chirlești earth flow are currently in progress, using the r.avaflow mass flow simulation tool (https://www.landslidemodels.org/r.avaflow/).

We developed a permafrost extension model fine-tuned for the Carpathians. BTS data was considered suitable for this purpose as it is the most widely spatially distributed across the Southern Carpathians. In March 2021 and 2022 we made additional field campaign measurements aiming to cover the missing environmental conditions like southern slopes, all landcover types or the new hot spots study sites of the project, i.e. Galeșu rock glacier. After filtering, we obtained a data base of 883 BTS points which we used to 8 machine learning algorithms, i.e. AdaBoost, Gradinet Boosting, Support Vector Machine, MaxEnt, K-neighbour, Logistic regression, Neural Network and Random Forest.

In order to train the models, the BTS dataset was split between a training and testing subset with a ratio of 70/30 % respectively, randomly sampled. We tested each algorithm and produced an ensemble map of permafrost distribution by intersecting all the obtained rasters in order to minimize each model’s weaknesses. The models were tested against independent measurements and the resulting maps can be interpreted in terms of permafrost presence probability with respect to the following predictors: altitude, slope, aspect, curvature, NDVI and snow persistence map (Fig. 5). The study areas include Retezat, Făgăraș, Parâng, Iezer, Țarcu and Godeanu massifs from the Southern Carpathians, the only ones that are considered to have a chance of hosting permafrost.

Figure 5 – Example of maps with environmental predictors (independent variables used against BTS values as dependent variable) used for permafrost distribution modelling: elevation, PISR, slope, slope curvature, NDVI and snow probability

The physically-based permafrost model CryoGrid was set up for steep rockwalls sites in the Southern Carpathians (based on classes developed in Schmidt et al. 2021, https://doi.org/10.5194/tc-15-2491-2021) and is planned to be used on rock glaciers as well, by adapting the model setup used by Renete et al. (2022, https://doi.org/10.5194/esurf-2022-39).

 

WP3 (Danube delta and Lower Danube floodplain)

SAR interferometry, geophysical surveys, cores and drills in Lower Danube Valley and Danube delta and their response to sea-level rise during the Holocene

Stratigraphy of the Danube floodplain and delta was mainly investigated by direct means (coring) but supplemented by geophysics especially by ERT profiles (Isaccea, Razelm, Enisala, Histria) which allow discriminating the contact between different contrasting sedimentary units (e.g. peat-sand-clays) or between sediments and the underlying bedrock. In the 2022 summer, new geophysics (using seismic methods combined with ERT) were used for the identification of former paleo-channels in the Danube floodplain.

Sediment sampling by cores and drills included 12 long cores (6-18 m) performed with percussion corer system in several locations from southern Danube delta, lower Danube floodplain and the lower reaches of Black Sea tributaries (Taița, Slava, Casimcea, Hamangia and Nuntași) to record several sea-level indicators necessary to reconstruct the Black Sea level oscillation and landscape changes during Holocene.

Additionally, in the spring of 2022, we succeeded to get two deep drills (up to 25 m) on the Lower Danube Floodplain on the sites of former Crapina and Jijila Lakes (drained in 1970s) in order to capture the full sequence stratigraphy covering Last Termination (22 ka – present) from Last Glacial Maximum till Anthropocene. Presently, the core samples are processed in different labs at PP (sedimentological, paleo-ecological and magnetical parameters) but also P2 (iTREX XRF scanning) and P3 (AMS dating; C/N measurements).

Based on the existing (Vespremeanu-Stroe et al., 2017) and the new cores and measurements got in the project we started to refine the Danube delta coastline mobility and subsidence processes in the last 8000 yrs.

10 short cores (0.7-1 m) from Danube Delta were extracted with a van Beeker system to get undisturbed sediments for ¹³⁷Cs and ²¹⁰Pb measurements used to estimate the recent sedimentation in delta plain. In this purpose, our strategy involves several profiles and locations from different units – fluvial banks and levees, lakes and lagoons, marshlands (incl. reed-marshes), floating islands (plauri) – which we expect to get different responses to sea-level rise in relation to sediment flux and local vertical movements. ¹³⁷Cs and ²¹⁰Pb measurements are in progress to P3 facilities from Bucharest and Slănic saline labs using 1-cm slices of sediments from the cores.

We have created a common database comprising ground displacement data from both in situ GPS/GNSS observations (from INCDFP GLASS and GEOECOMAR GeoPontica) between 2013 and 2021 and SAR interferometry analysis based on ESA Sentinel-1 mission between 2015 and 2019 (from The Synthetic Aperture Radar Persistent Scatterers Online Software Tool – PSTool developed by TerraSigna).

Figure 6 – Yearly deformation rates for different areas in the Danube Delta:  A) Sulina; B) Sfântu Gheorghe; C) Caraorman; D) Periteașca

Physical analysis of fluvial and deltaic sediments

Describing cores stratigraphy based on grain-size and sorting analysis; magnetic susceptibility measurements; XRF; organic matter content; elemental analysis as proxies for sediment texture and sedimentation environment was applied for each core (see above) and on the two deep drills following the lab detailed description and sampling. Grain-size is performed with a LSDPA Horiba-LA950 after the samples were treated with acetic acid (CH3COOH, 30%) and hydrogen peroxide (H2O2, 10%) to remove all carbonate and organic matter, for LOI the samples are burned in 3 steps at different temperatures and weighed after each step, for magnetic susceptibility is used a Bartington sensor (MS2C) with measurements at every 2 cm whilst for sediment geochemistry XRF Bruker Tracer S1 Titan device. Dozens of samples are in progress for AMS dating at P3. Moreover, on some of the cores (especially on the short cores sampled in Danube delta for assessing vertical accretion of the delta plain) with make additional measurements at P3: TOC, density, geochemistry.

Until now, in 2022, on the long cores from Lower Danube Floodplain, we add more new ages and data of TOC, LOI, and magnetic susceptibility that help to decipher two units with different sedimentation patterns:

The first one, which develops between ca 7500 and 5500 yrs BP, has a high sedimentation rate (2,5 mm/yr) and a small organic matter (7%) and clay content (<5%), in the same time has many thin layers of sand and silt. We interpret this unit as belonging to a dynamic floodplain with many avulsions and crevasse splays that try to keep up with the increase of relative sea level.

The second one is more homogeneous, composed of clay and peat, that develop in a natural regime, between 5500 years BP and 19^th century and has a smaller sedimentation rate (of 0,7 mm/yr), high organic matter (12%) and clay (10%) content. We interpret this unit as a more stable alluvial environment, a poor-drained floodplain with very few avulsions, in fact, many of the big lakes or swamps are formed at the start of this period (Brateş lake – 5300 BP; Crapina Lake – 5700 BP; Parcheş Lake – 5.700 BP; Fig. 7).

Palaeo-ecological analyses

Pollen analyses together with ostracoda and micro and macro-charcoal were done on three cores from the Lower Danube floodplain. Sediment samples of 1 cm³ were taken at 10 cm intervals and processed according to the standards. At least 300 arboreal pollen grains per sample were counted were possible; Lycopodium tablets were added to each sample to determine pollen concentration.

Palaeo-ecological analyses (mainly pollen and ostracoda) are in progress on different cores, most of them related to paleo-environmental reconstruction with or without connection to past human imprints. A new core from Jijila Lake was sampled for pollen and charcoal (macro and micro) analyses at an interval of 35 cm.  Three new 14C AMS dates were obtained that enable an improved chronological model. Results from macrochacoal analysis are presented, while we have already started the pollen and microcharcoal samples preparation. At this stage, the new macro-charcoal data together with grain-size and LOI analyses were corroborated with available archaeological data to develop a picture of the fire history in this area since Eneolithic. First evidence of possible anthropogenic fire activity is recorded in the Eneolithic, around 5100 BC. An important increase in fire signals is visible during the Gumelniţa and Cernavodă cultures (4500 – 3500 BC), followed by a decrease in in the Bronze-Age. Maximum fire event is recorded in the Roman period (~30% of the charcoal particles are from this layer). Other small peaks are from Byzantine and Mongol periods. These data are in a good relationship with the data from Oltina lake (Feurdean, 2020) and we hope that will be confirmed by the pollen and microcharcoal results.

……………………………………………..

 

Month 18 Interim report 2

WP2 Mountain landforms response to past and present climate change and future impact to hazard-related processes

Remote sensing and in-situ topography in alpine and landslides sites

Existing geomorphological maps (Urdea, 2000), landforms inventories (Onaca et al. 2017; Șerban et al. 2020), geomorphic field mapping and satellite images were used to create GIS-inventories of periglacial features in the central part of the Retezat Mts. Several tens of high-altitude moraines were mapped in Făgăraș Mts using satellite and aerial images as a preliminary step for understanding the deglaciation chronology.

 

 

 

 

 

 

 

 

 

 

 

 

 

In-situ ground-based DGPS measurements were conducted in Retezat (Galeșu, Judele and Căldarea Berbecilor rock glaciers: RG) and Făgăraș Mts (Doamnei RG) using a Topcon Hiper V high accuracy differential GPS.

Sentinel-1 InSAR and ALOS-PALSAR image pairs from 2016 to 2021 have been processed in SNAP and GAMMA software with different temporal baselines. Signals show clear displacement in RG in other areas in Romania with higher displacement rates, which confirms that the method is suited for the purpose. Processed ALOS-PALSAR data does not show any robust displacement signals, likely due to the lower spatial resolution and the reduced sensitivity to motion in the L-band.

Reference Digital Elevation Models were generated from drone-based photogrammetry and terrestrial laser scanner of Galeșu, Doamnei rockwalls-RG systems, Siriu dam debris-flow and Bâsca landslide. These models will be baselines for future terrain movements detection.

UAV-derived DEMs using structure-from-motion were obtained for Galeșu, Pietrele, Valea Rea, Bâlea and Capra RGs, for detailed morphology analysis by multi-scale topographic position analysis.

Geochronology of landforms and dynamics reconstruction

For relative dating of glacial and periglacial landforms, we performed extensive Schmidt hammer measurements on 8 RG in Retezat and 5 RG in Făgăraș, where samples were also collected for absolute age dating. The measurements were performed mostly on the terminal fronts and interior crests of complex RG. On each of the 30 sites, we tested 50 blocks with 2 or 3 shots. A total of 2289 shots were performed on 19 sites and 950 blocks in Retezat and a total of 1275 shots were performed on 11 sites and 569 blocks in Făgăraș.

During 2020-21, a total of 130 samples were collected for absolute age determination by Terrestrial Cosmogenic Nuclides (¹⁰Be, ³⁶Cl) surface exposure dating from glacial and periglacial sites covering 14 glacial valleys/cirques (76 samples) and 12 RG (40 samples) in Retezat and Făgăraș Mts. Depletion areas of 4 deep-seated landslides in the Curvature Carpathians were sampled, including inheritance reference sites for each (14 samples).

54 samples are in advanced chemical processing for ¹⁰Be and ²⁶Al dating (stage: cation exchange column) and 5 BeO and 5 Al₂O₃ targets are prepared for AMS measurements.

Samples were crushed, sieved, magnetic separated. The other minerals were eliminated with mixtures of HCl + H₂SiF₆. Then atmospheric 10Be was eliminated by HF (48%) dissolutions. Subsequently samples precipitated with NH₃ before successive separations through an anion exchange column and a cation exchange column.

Electrical resistivity tomography (ERT) and Refraction Seismic (RS) were conducted at two sites in the Retezat (Galeșu RG) and Făgăraș (Doamnei RG) in Sept 2021. The aim was to investigate permafrost occurrence and the internal structure of rock glaciers and talus slopes.

Two sediment cores of 4.9 and 6 m were sampled into a debris flow cone near Buda glacial Lake (2070 m, Făgăraș) using a percussion drilling set. Data will be used to characterize the past activity of debris flow and to indirectly infer the Holocene climate variability.

Climatic data analysis

We created a common database with all the ground surface temperature measurements in the Southern Carpathians. 25 GST sites were analyzed. Temperature-derived parameters were calculated: winter equilibrium temperatures (WEqT), mean annual ground surface temperature (MAGST), ground freezing and thawing index (GFI, GTI), snow cover duration (SCD), zero curtain interval (ZCI) and mean daily maximum temperature (Tmax).

Daily climatic data (temperature, precipitation and snow cover) from high altitude meteorological stations in the Carpathians (Vf. Omu – 2504 m, Vf. Țarcu – 2180 m, Bâlea Lac – 2038 m) were used to calculate the indices relevant for periglacial processes (mean monthly/annual air temperature, freezing and thawing indices, freezing/thawing cycles/yr, freezing days/month/yr, consecutive days with mean negative temperature, mean monthly/snow cover thickness, snow cover days/yr, precipitation/month, snowfall days/month/yr).

Development of a downscaling toolbox in Python was made for mountainous regions using global reanalysis models such as ERA5 ranging from 1950 to today. The toolbox allows for downscaling meteorological state variables as well as energy balance components at specific location or over an entire region.

Complex modelling for hazard assessment

For landslide susceptibility models for hazard scenario development at regional scale we have built the computer code and mathematical relations for landslide susceptibility models and numeric simulations and collected in situ data for validation. Digital elevation models were obtained for two study sites and in-situ temperature and precipitation monitoring started.

We developed a permafrost extension model fine-tuned for the Carpathians, based on multiple methods (random forest, decision tree, regression, etc.) using ground temperature data from BTS and GST measurements and geophysical data. Permafrost mapping for the Carpathians was performed with a statistical model trained by observations. As a first step, Machine learning permafrost distribution modelling based on in-situ ground temperature was carried for the mountain ranges Retezat, Făgăraș, Parâng and Iezer. The analysis includes an ensemble of models for uncertainty estimate. For permafrost distribution evolution models computing we prepared a data set with meteorological parameters, including temperature and precipitation data, downscaled from ERA5, ranging from 1950 to 2020. At the same time, the physically-based permafrost model CryoGrid was set up for sites in the Carpathians, including steep rockwalls.

WP3 Response of Danube Delta and its coast to sea level rise, increasing ex­tremes (storm surges, river floods) and human pressure

SAR interferometry, geophysical surveys, cores and drills in Lower Danube Valley and deltaic plain

Stratigraphy of the Danube floodplain and delta was mainly investigated by direct means (coring) but supplemented by geophysics especially by ERT profiles (Isaccea, Razelm, Enisala, Histria) which allow discriminating the contact between different contrasting sedimentary units (e.g. peat-sand-clays) or between sediments and the underlying bedrock. Sediment sampling by cores and drills included 12 long cores (6-18 m) performed with percussion corer system in several locations from southern Danube delta, lower Danube floodplain and the lower reaches of Black Sea tributaries (Taița, Slava, Casimcea, Hamangia and Nuntași) to record several sea-level indicators necessary to reconstruct the Black Sea level oscillation and landscape changes during Holocene. Based on the existing (Vespremeanu-Stroe et al., 2017) and the new cores and measurements got in the project we started to refine the Danube delta coastline mobility and subsidence processes in the last 8000 yrs.

10 short cores (0.7-1 m) from Danube Delta were extracted with a van Beeker system to get undisturbed sediments for 137Cs and 210Pb measurements used to estimate the recent sedimentation in delta plain. In this purpose, our strategy involve several profiles and locations from different units – fluvial banks and levees, lakes and lagoons, marshlands (incl. reed-marshes), floating islands (plauri) – which we expect to get different responses to sea-level rise in relation to sediment flux and local vertical movements. 137Cs and 210Pb measurements are in progress to P3 facilities from Bucharest and Slănic saline labs using 1-cm slices of sediments from the cores.

We have created a common database comprising ground displacement data from both in situ GPS/GNSS observations (from INCDFP GLASS and GEOECOMAR GeoPontica) between 2013 and 2021 and SAR interferometry analysis based on ESA Sentinel-1 mission between 2015 and 2019 (from The Synthetic Aperture Radar Persistent Scatterers Online Software Tool – PSTool developed by TerraSigna). These data will be used in the next phase (2022) to extract the present subsidence rates and to analyze their different spatial distribution throughout Danube Delta.

Physical analysis of fluvial and deltaic sediments
Describing cores stratigraphy based on grain-size and sorting analysis; magnetic susceptibility measurements; XRF; organic matter content; elemental analysis as proxies for sediment texture and sedimentation environment was applied for each core (see above) following the lab detailed description and sampling. Grain-size is performed with a LSDPA Horiba-LA950 after the organic matter is removed with CH3COOH and H2O2, for LOI the samples are burned in 3 steps at different temperatures and weighed, for magnetic susceptibility is used a Bartington sensor (MS2C) whilst for sediment geochemistry XRF Bruker Tracer S1 Titan device. Dozens of samples are in progress for AMS dating at P3.

Palaeo-ecological analyses
Pollen analyses together with ostracoda and micro and macro-charcoal were performed on 3 cores from the Lower Danube floodplain. Sediment samples of 1 cm3 were taken at ca. 10 cm intervals and processed according to the standards. At least 300 arboreal pollen grains per sample were counted were possible; Lycopodium tablets were added to each sample to determine pollen concentration.

Microscopic and macro-charcoal particles were also counted as a proxy of the fire regime and human pressure history.

OVERVIEW OF THE PROGRESS OF WORK TOWARDS THE OBJECTIVES OF THE PROJECT

O1. Reconstruction of past landscape changes in the Carpathians and the lower Danube driven by Holocene climate change and sea-level oscillations
The new absolute exposure ages indicate intense rockwall degradation via rock-slope failures took place in the Carpathians in Early Holocene, reaching the highest magnitudes 11.6 – 9 ka ago especially above 1800 m. We associate the present distribution of RW with this periglacially-active period which was the last interval when the rock surfaces were substantially active. Also, RG surface ages place their formation largely during the same period. Complementary, the cores obtained from Buda Lake (being processed at UB) are expected to provide relevant data on mid-late Holocene debris-flow activity in respect to paleo-climate. A new core is expected to be obtained in 2022 in the alpine area (Bâlea Lake).

Rock glaciers age and evolution is in progress by applying both absolute and relative age dating on several RG in Retezat and Făgăraș Mts. Rebound values (Schmidt hammer) increase with altitude confirming the vegetation pattern (from forest- to lichen-covered RG) indicating younger ages with altitude increase. Dozens of rock material were sampled from RG fronts on altitudinal transect but most of them are still being processed. Results from Doamnei RG (2050-2150 m asl, Făgăraș) indicate a Younger Dryas age of formation. Taking into consideration the altitudinal range of the RG distribution from Southern Carpathians (1600 – 2300 m asl) we expect Late Glacial ages for most of them and probably early Holcoene ages for the highest ones.
For the upper Danube delta evolution we used the sedimentary, pollen, and ostracods from a record (Somova Lake) that covers the last 7500 yrs. The new data reconstruct the evolution of the upper fluvial delta, revealing the conversion from a palaeo-channel to a shallow delta plane lake, prior to 5700 BP and to a lacustrine reed marsh, thereafter. Pollen data reveal a much more forested landscape than today and supports the early presence of Carpinus betulus and Fagus in comparison to the uplands. After 5700 BP the pollen spectra show a transition from a more forested to an open xerothermic landscape. The tree stands were dominated by Quercus along with Carpinus, Betula, Ulmus, and Tilia. Important alteration of the natural environment in the region is visible starting from 3200 BP (late Bronze Age) through deforestation and conversion to agropastoral land, resembling its modern ecological character.

The multi-proxy analysis (grain-size, LOI, magnetic susceptibility, meiofauna, pollen, AMS dating) performed along an array of 10 cores from Lower Danube Floodplain reveals a two stages pattern of evolution for the last 8000 years that is in a close connection with the evolution of sea level: (i) a highly dynamic well-drained floodplain (8-5.5 kyr BP), which is marked by successive transitions between wet and dry conditions, evidenced in coarse and fine-grained intertwined sedimentary units in a rapid sea-level rise stage (6-3 mm/yr) and (ii) a meta-stable poor-drained floodplain for the last 5.5 kyr which is marked by a presence of numerous lakes and wetlands evidenced in fine-grained sediments and peats units.

O2. Recent to present response of landforms and geo-hazards to climate trends and increasing variability

Preliminary modelling exercises with Delft3D for long-term delta lobe evolution were established. A major breakthrough is achieved in this project in modelling, for the first time, of complex asymmetric wave influenced delta lobes. Wave influenced delta lobes show a variety of morphologies related
mainly to their degree of protrusion or deflection of the main river channel. Using different balances of river sediment input (Qs) and longshore wave driven sediment transport, the model was able to simulate the different morphologies of wave-influenced lobes from the Danube Delta. Generally, concave shoreline exists updrift (north) of the river mouth and convex downdrift (south) of the river mouth. 30 preliminary simulations were run with Delft3D for the purpose of this project, and the most relevant were selected. Future model set-up will be calibrated long-term shoreface and delta-front evolution data from the Sf. Gheorghe lobe.

Danube Delta coastline experienced variable dynamics in the last 160 years, depending on the factors and processes which had the leading role at different temporal and spatial scales. At centennial timescales, shoreline evolution was highly influenced by the threefold decrease of Danube sediment discharge in the last century (mainly after 1950, as a result of dams construction in the Danube watershed), especially along the accumulative sectors of secondary deltas.
On multi-decadal scale, storminess (because of climate variability) was the driving factor for coastline changes. Human influence superposed on the natural response of the coast to variability of climatic, hydrological and hydrodynamic (marine) factors. All observed changes resulted in different shoreline morphology and configuration along Danube Delta coast (e.g. the transition from fluvial- to wave-influenced morphology of the Chilia lobe; the shifting of the Sf. Gheorghe mouth from an asymmetric to a deflected wave-influenced delta morphology). For the last three decades, we identified an extension of erosional processes.

Climate variability trends in the alpine area
Meteorological observation from long term weather stations have been acquired over the Retezat and Făgăraș Mts. Data will be compared to downscaled reanalysis time series for validation and correction on the specific topography of periglacial landforms in the Carpathians. Downscaled time series and observation will be the baseline for trend analysis and generation of climatic forcing dataset for permafrost and snow models. Model selection includes Cryogrid and FSM, used with the purpose to map and study permafrost in the Carpathians. To get data for the model, UiO implemented downscaling tool TopoScale, which may use coarse-scale reanalysis data available since the 1950s in the Carpathians. The analysis is finalised while writing of a publication is in progress.
Climate data analysis conducted so far shows that the alpine belt of the Southern Carpathians is facing significant warming, which is reflected by the substantial increase of MAAT, thawing degree days and freezing degree days values, especially after 2000. The number of days with mean negative temperatures indicates a strong decline in the last decade at all three stations. Air temperature rise is enhanced during the summer season. Positive trends of MAT reach maximum at Bâlea Lac (+0.54°C/decade) and are similar at Țarcu (+0.28°C/decade) and Vf. Omu (+0.27°C/decade).

Precipitation does not show a clear trend in the alpine region of the Southern Carpathians with an increase at Bâlea Lac (166.7 mm/decade) and Țarcu (6.96 mm/decade) and decrease at Vf. Omu (47.9 mm/decade). Conversely, a marked decrease of the heavy precipitation (>20 mm) frequency was recorded in 1991-2020 compared with 1961-1990 at Vf. Omu (-26%) and Vf. Țarcu (-15%).
GST data analysis from openwork structures in RG and talus slopes were combined from P2 (19 locations) and CO (6 locations) in Retezat, Făgăraș and Iezer Mts. The analysis revealed that permafrost is likely at most of the investigated sites since WeQT values < -2°C at the end of the winter are recorded regularly. The main factors controlling WeQT are the air temperatures at the beginning of the winter season (before the onset of the insulating snow cover) and the date of the thick snow cover onset. At the sites monitored by CO, MAGST registered increasing trends at four locations from a total of six. The BTS period varied multi-annually but with no significant trend maintaining into the permafrost possible or probable thresholds in almost all years.

Recent and present alpine landforms response to climate trends
Snow avalanches on 17 avalanche paths located in the Făgăraș Mts were reconstructed for the last century based on 956 tree-ring data. Rockfall and debris flows in two sites in the Făgăraș Mts were analysed and dated, based on samples extracted in 2021. Landslide reactivations in the Curvature Carpathians were dated based on samples extracted in 2021.

Currently, we look for the relationship between major avalanches (1988, 1997 and 2005) and climatic variables and trends. Similarly, work regarding the link between rockfall, debris flows, landslide reactivations and climatic triggers is in progress. Significant values for rockfall frequency have been obtained for 1963, 1970, 1973, 1990 and recent: 2003, 2006, 2008, 2018.
Present displacements have been addressed by Sentinel-1 InSAR and ALOS-PALSAR image pairs processing from 2016 to 2021. Displacements rates in the Făgăraș are of ca. few cm/yr, making it difficult to distinguish displacement from noise but a robust signal shows movement in Doamnei RG where there also have been observed indications of displacement during fieldwork, and where ice/permafrost is suggested through geophysical soundings.

*This section is updated in agreement with the project implementation calendar and results achievement.

Working together for a green, competitive and inclusive Europe