WICLAP - wpływ zmian warunków zimowych i jakości powietrza na lasy

The work within work package 2 was concentrated on analysis of meteorological conditions at the selected test sites in order to identify temperature and precipitation anomalies and tendencies and to make forecasts of climatic situation based on the identified trends. In case of Arctic study area a long series of meteorological data – temperature and precipitation -  were acquired, covering over 90-year’s period (1924 – 2016).  The original data from weather stations were analyzed using various winter warm event indices (number of warm days, number of melt days, positive degree-day sum, number of melt and precipitation days, total winter precipitation amounts) and applying appropriate thresholds for warming event identification.  In order to study trends the linear regressions of the indices were analyzed over the past 15, 30, 50 and 90 years, as well as for the future 50 – 100 years. As a result of these analyses the following conclusions characterizing meteorological conditions in the past and in future were drawn:

• Weak positive trends in winter warming events are found over the past 90 years. This is due to the warm winters of the 1920s and 1930s, the period most similar to the present (2000–14) in terms of winter climate. Midwinter warming events were more frequent during the 1920s and 1930s, similarly to the present. However, early and late winter have a clearer increasing trend over the past 90 years.

• For the past 50 years, we find strong increasing trends in occurrence and intensity for all climate indices, all statistically significant. The main pattern follows the general temperature development of arctic areas. The trends are stronger to the north, with the largest trends at the arctic islands. Here, the rate of melt days  in winter increased by up to 9.2 days decade 21. For melt days with precipitation (mainly as rain) the increase in precipitation sum is up to 20.8 mm decade21.

• During the past 15 winters (2000–14), warming events have become more frequent and intense, seen in a 90-yr perspective. Of the top five extreme winter warming years, 37% occurred within the most recent 15-yr period, followed by the 15-yr period from 1925 to 1939, which had 20% of the top five extreme winter warming years.

• Regional climate model simulations over the next 50–100 years indicate that the strong past 50-yr trends will continue. In northern Scandinavia, simulations indicate a doubling in the number of warming events while the arctic islands may see a threefold increase in the number compared to the 1985–2014 reference period.

The results of analysis in graphic form are presented in figure 1.

Fig. 1. Results of analysis related to appearance of winter warm days in Near Arctic Region

These results are valid for stations in the Near Arctic Region and hence obtained from a dataset of limited spatial coverage. To evaluate the representativeness of this analysis in a broader geographical perspective, it is recommended that a similar study is carried out using reanalysis data for the entire pan-arctic region, elucidating possible variability in extreme events occurring in continental and maritime climates of the arctic region. We also suggest carrying out this study with a larger ensemble of regionally downscaled projections covering the Arctic (e.g. CORDEXArctic). A larger spread of simulations will provide even higher trust to the findings of future changes. The findings in this study have improved our understanding of the regional climate variability in the Near Arctic Region. The information is important for predicting the impact of climate change in the arctic environments, as well as for developing adaptation strategies to deal with these changes.

Polish results:
The activity within work package 3 was divided into two parts, devoted to analysis of low-resolution satellite images and high-resolution satellite data. At first part of the work the activity was aimed at evaluation of applicability of low-resolution satellite data for studying various environmental and climatic conditions in Polish forests. The results of the work revealed, that both aspects of forest variability can be to a large extent monitored with the use of vegetation indices derived from NOAA AVHRR images. Forest areas located in various climatic regions – under impact of continental climate in northeastern Poland and under impact of maritime climate in southwestern Poland are characterized by different NDVI curves, especially at the beginning and end of growing season. General species composition within the study areas – dominance of coniferous or deciduous / mixed forests – also has visible impact of NDVI levels. In case of high temperatures at the end of winter and beginning of spring (2014 case) NDVI for forest study areas located in southern Poland begins from lower values due to later start of growing season in the Beskid and Karkonosze mountains. It is next compensated by favourable temperature conditions, reaching in May – June similar values to the remaining forest areas. Forests in mountainous regions are more sensitive to NDVI fluctuations due to more variable climatic conditions. The lowest NDVI values in the mid-growing season are observed for Augustowska and Knyszynska Forests, which are the study areas more influenced by continental climate than the remaining ones.

Climatic situation has also an impact of values of vegetation index. In case of the the “cold” year, represented by low temperatures at the end of winter / beginning of spring (2006 case). In this case NDVI values for all study areas are low at the beginning of growing season (the lowest one for Karkonosze test site), reaching quite similar values in the mid-growing season. At the second part of the growing season (July – September) test areas represented predominantly by mixed forests (Bialowieska and Borecka Forests) reveal higher NDVI values then the remaining ones. Again, lower NDVI values can be observed for Augustowska and Knyszynska Forests, influenced by more continental climate. Comparison of NDVI values depending on year (2006 – “cold” year, 2014 – “warm” year) is presented for northeastern forests in fig. 1.

Fig. 1. Comparison of NDVI changes for northeastern forests – 2006 and 2014

Results of analysis of NDVI index based on low-resolution satellite data were supported with study of the second index, which characterizes forest condition – Vegetation Condition Index – VCI. The results of its analysis in 2000 – 2016 period for all study areas point out the similar relations as in case of NDVI index, revealing much lower VCI values at first part of vegetation period (April – May), while comparing “cold” and “warm” years, which are compensated values at the later stage, of plant development.

Trend analysis has been done on the basis of vegetation indices derived from low-resolution NOAA AVHRR and MODIS data covering 2000 – 2016 period. The results of this analysis led to the conclusion, that there is no significant change of NDVI within the reporting period at the beginning and at peak of the growing season, which means that within last 16 years no statistically significant impact due to climatic variations was observed. This conclusion is in agreement with the analysis of climatic data in the same period – air temperature and hydrothermal index characterizing drought conditions. Trend analysis of these parameters at the beginning of growing season and at peak of growing season did not reveal statistically significant changes through the reporting period 2000 – 2016.

The second part of the works within work package 3 was devoted to analysis of high-resolution satellite data for studying environmental and climatic conditions in Polish forests. Three types of high-resolution images were used: Landsat images, SPOT 5 images and Sentinel 2 images (also WorldView 2 image of one study area was used for comparative purposes). The analysis was concentrated on northeastern forests, where sufficient number of images could be obtained for vegetation periods. Several vegetation indices were derived from high-resolution Landsat and SPOT images:
- Normalized Difference Vegetation Index – NDVI
- Enhanced Normalized Difference Vegetation Index – ENDVI
- Enhanced Vegetation Index – EVI
- Normalized Difference Infrared Index –NDII
- Ratio Drought Index - RDI
- Disease Stress Water Index – DSWI

Particular indices were analyzed in a temporal profile through the whole vegetation seasons for three study areas: Bialowieska Forest, Knyszynska Forest and Borecka Forest. Results of these analyses reveal usability of Disease Stress Water Index – DSWI and Normalized Difference Infrared Index - NDII for differentiating some tree species within particular forest areas.

At the second stage of the works impact of variability of environmental conditions on values of vegetation indices derived from high-resolution satellite data was investigated. Two components of forest environment were studied: type of forest site and stand mixture. Three main types of forest site were considered: humid site, fresh site and dry site. Also three levels of stand mixture were taken into account: pure stands, mixed stands with 70 – 90 % of dominant species, mixed stands with 50 – 70 % of dominant species. The analysis of various vegetation indices revealed that DSWI proved to be the most sensitive to differentiating both for forest sites and stand mixture. In case of applying Sentinel 2 data additional vegetation indices, which are based on narrow spectral red-edge bands, were calculated. Two indices dedicated to Sentinel 2 proved to be the most useful for differentiating forest features: tree species, forest site and stand mixture – Triangular Vegetation Index – TVI and Transformed Chlorophyll Absorption Ratio Index – TCARI. The relationships between vegetation indices derived from high-resolution satellite data and meteorological conditions within the study areas are presented in the description of work package 7 – task 7.2.

Norwegian results:
The activity within work package 3 was divided into three parts, devoted to analysis of low-resolution satellite images, high-resolution satellite data and aerial imagery. In case of low-resolution satellite data for Svalbard study area two types of images were applied: 1-km NOAA AVHRR images (30-year’s dataset) and MODIS NDVI 8-day product (2000 – 2014). In addition, local meteorological data were available to link maximum NDVI values to the temporal trends of the mean growing season (summer) temperature for the study area. Over this 30-year period covered by NOAA AVHRR data there were positive linear trends of 0.0017 yr-1 for NDVImax (average increase of 29 %, p < 0.05) and 0.065 °C yr-1 for mean summer temperature (59 %, p< 0.01), respectively. NDVImax and mean summer temperature were also positively correlated.

The years 2005-2007, 2010, 2012 and 2014-15 had NDVImax much lower than average (AVHRR). Low NDVImax is an indicator of plant stress. These low index values coincide with reports of ground-icing episodes and/or winter warming events (Vickers et al. 2016 and references therein). The reduced NDVI values in the peak season of the above mentioned years were partly confirmed by analysis of MODIS NDVI images.

In case of high-resolution satellite data four types of images were applied RapidEye 2014, Formosat 2011-2012, WorldView 2 2013 and Sentinel 2 images from 2016. Analysis of 5-m RapidEye images acquired over Longyearbyen in the same period as the spectrometer measurements were done, showed positive results. A regression analysis of the in situ-derived RENDVI (ASD Fieldspec) of the species Dryas octopetala and the satellite extracted RENDVI (RapidEye) for the different sites dominated by this species showed a strong relationship of R2 = 0.82, p=0.036. This shows that it is possible to upscale from field-based assessments of vegetation stress (both pollution and climatic stress) to local and regional level Assessments of 8-m data from the Formosat satellite acquired in 2011 and 2012 reveal that the winter warming event (ground icing) in 2012, which caused detrimental damage to the dwarf shrub vegetation, can be detected on the 2012-image. Also assessments of digital orthophoto and WV2-data from Ny-Ålesund reveal significant reduction of NDVI in the period 2009 to 2013, due to the large damage on the dwarf shrub vegetation during the winter 2011-2012. In case of Sentinel 2 data only two cloud-free images were acquired for Svalbard Longyearbyen test site, so their application was limited, nevertheless quite good relationship was obtained between NDVI extracted from S-2 data and that measured during field campaigns (r2 = 0.46).

In addition, applicability of aerial images acquired with the use of UAS (Unnmaned Aircraft System) was studied. The analysis of NDVI’s for four plant communities acquired by UAS from 100 m level revealed, that they correlate well with NDVI’s obtained from ground measurements (community level) – the correlation coefficient was R2 = 0.75 (p<0,01).

Work Package 4 covered field data acquisition and analysis conducted on research areas in Poland throughout the period of 2014 - 2016. The acquisition of the data was carried in north-eastern (NE) and south-western (SE) parts of Poland - lowland and mountainous areas accordingly. A total of 53 test homogenous test polygons of a southern sun exposure were selected in the following geographical areas of Poland: Bialowieza forest (NE, 8 polygons), Knyszynska forest (NE, 13 polygons), Borecka forest (NE, 8 polygons), Karkonosze Mountains (SW, 12 polygons), Beskid Zywiecki Mountains (SW, 12 polygons). Dominant deciduous and coniferous tree species were selected as the object of analysis: birch (Betula pubescens, NE), oak (Quercus robur, NE), hornbeam (Carpinus betulus, NE), alder (Alnus glutinosa, NE), beech (Fagus sylvatica, SW), spruce (Picea abies, SW, NE), pine (Pinus sylvestris, NE), fir (Abies alba, SW).

The following information were acquired:

• plants' spectral characteristics with ASD FieldSpec 3 (in 2014) and 4 (in 2015, 2016) spectroradiometer with PlantProbe collecting reflected electromagnetic radiation (range of 350-2500 nm);
• chlorophyll fluorescence values in non-adapted and dark-adapted states with Chlorophyll & fluorescence OS1p Chlorophyll Fluorometer;
• plants' surface and air temperatures with IRtec MiniRay 100 pyrometer and noncontact thermometer (only in 2014);
• chlorophyll, anthocyanins, flavonoids and nitrogen content with Dualex Scientific+ Polyphenol & Chlorophyll-Meter;
• coordinates of test polygons and specimens with Trimble GPS;
• documental photography (various cameras).

All data were gathered at the beginning (May) - I, during (June, July) - II and at the end (late August, September) - III of each year of data acquisition to portray phenological changes occurring in plants. Which in total gave 18 field measurements trips during 6 field campaigns: 3 trips every year for NE and SW Poland separately.

Collected data were analyzed statistically. Spectral characteristic curves were visualized and compared. Visual analysis of changes occurring between measurement periods (I, II, III) were made to identify wavelengths showing changes occurring in plants. The data were compared using ANOVA (analysis of variance). Statistical analysis allowed to select those wavelengths which showed changes in amount of reflected electromagnetic radiation during each measurement year by comparison the acquired data in pairs: I - II and II – III measurement periods.

Comparison of all acquired data showed how the amount of reflected electromagnetic radiation changed in each of measurement periods separately (I, II, III) between years of 2014- 2016.

In case of deciduous species comparison of beginnings (I measurement period) of the measurements years showed that the most prominent changes occurred in plants were observed in blue and red visible light, which may suggest variations in use of light by plants during processes of photosynthesis. However, this observation was not followed by any significant differences in green visible spectrum. Changes were also observed in red edge range of spectrum as well as whole range of near infra-red, which provides information on variations occurring in plants cell structures. Changes in mid infra-red were registered in wavelengths of 1400 – 1700 nm and 2000 – 2100 nm. Coniferous species data showed statistically significant changes in blue light of the visible part of the spectrum and minimal changes near 700 nm wavelengths. Close to none changes were registered in near infra-red, while considerably more changes than in deciduous species were noted in mid infra-red.

Differences occurring in plant’s spectral reflectance curves mid-year (II period) showed less changes in visible spectrum for deciduous species (no changes in red-edge wavelengths) while more for coniferous species (blue, green and red parts of the visible spectrum did show variations in reflectance values). Whole near-infra red spectral range was statistically significant for deciduous species, while no changes were observed for coniferous species. Opposite situation can be seen in case of mid infra-red: close to none changes were observed for deciduous species, while whole spectral range has showed differences between years for coniferous species.

Changes noted towards the end of every vegetative year (III) were observed for deciduous species only in visible blue spectral range, red edge, near infra-red close to 1275 – 1300 nm and mid infra-red 1850 – 1900 nm. For coniferous species approximately whole spectral range of 350 – 2500 nm proved to be significant.

The analysis of spectral response curves showed that the most significant changes between measurement years were observed in case of deciduous species at the beginning of the measurement years (I) while the least in the end (III) showing that the dynamics of growth of the species might have been different between years and that the starting point of plants’ development may have shifted depending on weather conditions or plant-specific factors. Lack of changes towards the end of the analyzed years shows that dynamics of plants development is quite similar and stabilizes with plant’s growth. Moreover, it may show that analyzed species did not reach yet point of senescence. Opposite situation noted for coniferous species may result either from plants’ specific character – evergreen plants – or variations in data acquisition due to plants structure.

Next spectral reflectance curves were used to calculate remote sensing vegetation condition indices. The indices focused on such aspects of the plants as: general condition and vigor of vegetation, pigment content (chlorophyll, carotenoids, anthocyanins), nitrogen content, water content, amount of dry mass, use of light in the process of photosynthesis and dynamics of senescence. The indices were statistically analyzed to delimit them only to those which showed statistically significant changes occurring between measurement periods I-II and II-III. The statistical analysis was based on ANOVA and Kruskal-Wallis ANOVA.

For deciduous species indices which proved to be statistically significant were general vigor broadband indices such as NDVI, SR, EVI, ARVI, SAVI and OSAVI, WRDVI. Most of narrowband greenness/red edge/chlorophyll dependent indices were significant as well (NDVI705, mSR705, mNDVI705, VOG1-3, NDPI, NPCI and SRPI, RENDVI, REPI). Out of protective pigment indices significant were ARI1-2 analyzing anthocyanins content and for carotenoids CRI1-2.; NDNI analyzing amount of nitrogen in a plant as well as NDLI analyzing dry material content. Amount of water can be depicted by NDWI, NDII or MSI, while use of light in the process of photosynthesis was best described by PRI.

For coniferous species statistically significant indices were: SR, VARI (general condition), mNDVI705, VOG 1-3, MCARI, TCARI, TVI (chlorophyll based narrowband condition indices), MRENDVI, REPI (red edge indices), NDLI (dry matter content), ARI2 (anthocyanins content), NDNI (nitrogen content), PRI (use of light in the photosynthesis) and NDII, NDWI, WBI (water content).

Bioradiometric data such as: chlorophyll, anthocyanins, flavonoids and nitrogen content as well fluorescence values were also analyzed statistically for significant changes. In 2014 the highest amount of changes was observed for nitrogen and anthocyanins content, in 2015 in nitrogen content and in 2016 in nitrogen and chlorophyll content. Fluorescence measurement did not show any trend and there were either no significant changes or the changes highly varied between species with no trend whatsoever.

Correlation of bioradiometric data with calculated remote sensing vegetation indices allowed to further investigate changes occurring in the plants. It showed:

• Chlorophyll content: for deciduous species strongest positive correlations were observed with chlorophyll content-based as well as red edge indices (NDVI 705, mSR 705, mNDVI 705, VOG 1, PSRI, Green NDVI, NDNI1, MRENDVI, MRESR, RENDVI,REPI) while negative correlations with water (WBI, NDWI, NDII), pigment (VOG 2, VOG 3, TCARI, TVI, VARI), nitrogen content (NDNI) and dry matter indices (NDLI). For coniferous species no correlations were observed.
• Flavonoid content: for deciduous species no positive correlations were registered, however strong negative correlations were observed with PRI, NDNI, NDLI, TCARI and TVI indices. For coniferous species positive with VOG 2, VOG 3, CAI, negative with: NDVI 705, mSR 705, mNDVI 705, VOG 1, PRI, Green NDVI, NDNI1, MRENDVI, MRESR, RENDVI, REPI.
• Anthocyanins content: for deciduous species positive correlations were observed with VOG2-3, NDII and TCARI and negative with PRI, PSRI, NDNI1. For coniferous species positive with MCARI, negative with PRI.
• Nitrogen content: for deciduous species correlations were exactly the same as for chlorophyll content. For coniferous species positive correlations were noted with WBI while negative with MSI.

Correlation of fluorescence values with remote sensing indices showed that:

• Deciduous species: in a non-adapted state positive correlations were noted for EVI, PRI, NDNI, CRI 1, CRI 2, TVI, while negative correlations with NDLI, CAI, ARI 2. In dark-adapted state positive correlations were observed with SIPI, PSRI, MSI, NDPI while negative with WBI, NDWI, NDII, SRPI, VARI. In case of deciduous species correlation of fluorescence with vegetation indices was visible mostly in case of light-related indices, protective pigment indices and water content.
• Coniferous species: non-adapted fluorescence values showed no correlations with any of the indices while dark-adapted values with nitrogen content index NDNI1, red edge REPI (positive correlations) and pigment indices TCARI and MCARI (negative correlations).

The analysis showed that all species show more phenological changes at the beginning of the vegetation period rather than towards its end. More changes can be observed in deciduous species, especially regarding chlorophyll content and related shift in red edge. Observations linked to high amounts of chlorophyll and strong significance of use of light in the process of photosynthesis along with a decrease in amount of dry matter suggest a grow of a plant. At the same time there are no strong changes visible in vegetation indices depicting phenology of coniferous species. Being mostly ever-green species, there were no data recorded which showed strong changes either in chlorophyll content or amount of dry matter. However, some changes may be observed towards the end of vegetative years, especially in amount of protective pigments such as carotenoids and anthocyanins. More significant role attributes to changes in water content, which may be connected to temperature and humidity fluctuations between measurement months and years. The use of hyperspectral radio- and bioradiometric data along with remote sensing vegetation indices allowed for a detailed analysis of changes in plants phenology and to assess which of the remote sensing vegetation indices may be used as a mean of a long-term plant condition monitoring.

Within Work Package 5, according to the plan, the nine biota sampling sessions in Polish territory were carried out. In Poland samples of epigeic moss Pleuroziom schreberi, epiphytic lichen Hypogymnia physodes, leafs of birch Betula L. (interchangeably: oak, hornbeam and alder) and needles of spruce Picea abies (L.) H.Karst. (interchangeably: pine and fir) were collected. In samples collected during nine sessions, concentrations of Mn, Ni, Cu, Zn, Cd, Hg and Pb were determined.

The collected samples were cleaned from mechanical impurities, dried at the temperature of 323 K and stored in tightly sealed polyethylene containers. Homogenised samples with the mass of 0.4 g were mineralised in the mix of nitric acid and hydrogen peroxide in  a microwave mineraliser MARS-X made by CEM. MERCK reagents were used to prepare the solutions. Concentrations of the following elements were determined in demineralised samples: Mn, Ni, Cu, Zn, Cd and Pb, with the FAAS method, using the iCE 3000 spectrometer made by Thermo Electron Corporation (USA).To determine mercury concentrations in the plant samples, the AMA 254 analyser was utilized.

Interlaboratory comparison of determination results was included in quality assurance procedures. The measurements were carried out in Opole University labs and in accredited lab in ZWP EMITOR (Opole, PL). Concentration of metals was also determined in reference materials BCR-414 plankton and BCR-482 lichen. The results of  heavy metals determination in reference materials and results of comparison studies indicate that the method related determination uncertainty does not exceed 20%.

To determine the areas subjected to deposition of heavy metals in the studied regions in Poland, the following evaluation criteria were considered:
1. A comparison of the areas studied in terms of the average content of the studied analytes in biota.
2. Assessment of seasonal increase in analytes concentrations in the tested biota parts.
3. To compare analyte concentrations in epigeic moss and in epiphytic lichen collected in Poland the comparison factor (CF) was calculated.

Field studies were preceded by an analysis of the literature data on the historical changes in the concentrations of heavy metals in biota samples collected on the areas studied.

The main conclusions from the research conducted in the years 2014-2016 in Poland:

1. The analysis of literature and the own research indicates that in the period 1975-2016 there was a significant improvement in the quality of the environment in regions covered by the study.
2. The results show a relatively high, compared to other samples, manganese accumulation in the leaves. The main source of increased manganese concentration in atmospheric aerosol is probably the soil dust. Big local background level precludes determination of manganese originating from distant sources.
3. Similar like for manganese, the highest nickel concentrations were determined in leaves. Also, the highest concentrations of nickel were found in samples of moss, leaves and needles collected in the Beskidy Mountains.
4. The concentrations of manganese and copper in the studied samples do not indicate significant differences with regard to seasonal changes and spatial distribution in the studied areas, which may indicate the soil as the main source of their origin.
5. In terms of contamination with Zn, Cd and Pb the studied areas can be ranked in the following order: Beskidy Mountains> Karkonosze Mountains> forests of north-eastern Poland.
6. With the exception of the Borecka Forest, the concentrations of mercury accumulated in the samples of moss and lichens in forests the north-eastern Poland are lower than the concentrations of mercury determined in samples collected from the area of the Beskidy and Karkonosze Mountains.
7. Seasonal changes in concentrations of Cd and Pb and Ni in the case of Knyszyńska Forest and Białowieska Forest, indicate the low emission during the heating season as the main source of contamination of the studied areas.
8. Distribution and seasonal changes in the concentrations of Zn and Hg indicate the distant emission or, as in the case of the Knyszyńska Forest, the territorial emission.
9. In the studied areas can be distinguished the sites with relatively high deposition of contaminants:
a) Karkonosze Mountains – site to the south-east of Szklarska Poreba,
b) Beskidy Mountains – site to the south of Sucha Beskidzka,
c) Borecka Forest – north-western part of the Forest,
d) Knyszyńska Forest – site under the influence of territorial emission from the city of Białystok,
e) Białowieska Forest– areas along the Hajnówka-Białowieża road.

In 2014 on Svalbard, near Longyearbyen, were collected samples of various species of mosses, Thamnolia vermicularis lichens, polar willow Salix polaris, Cassiope tetragona and soil. In the samples were determined the concentrations of: Mn, Ni, Cu, Zn, Cd, Pb and Hg and the activity concentrations of 40K, 137Cs, 210Pb, 212Pb, 212Bi, 214Pb, 228Ac, 231Th and 235U.The surface activity concentrations of gamma radioactive isotopes were determined, in situ. Identification of natural and artificial gamma radioactive isotopes enabled assessment of the impact of marine aerosol pollution in the studied area. Thesis on the pollutants movement in an easterly direction confirms the concentrations distribution of metals determined in biota and soil elements. Concentrations of Mn, Ni, Cu and Zn are largely higher than the median in samples taken east of Longyearbyen, compared with the western side. This may indicate a local delivery of these analytes from industrial plants deployed in Longyearbyen, but it can be also a result of different chemical soil compositions  in the east and west sides of Lonyearbyen.

The main conclusions from the study conducted in 2014 in Svalbard:
1. In the studied area one of the sources of anthropogenic contamination is the sea aerosol, as indicated by the linear correlation between the concentrations of: 137Cs, 210Pb, 231Th, Pb and Hg in the Surface layer of the soil. The concentrations of these analytes decrease with distance from the shoreline.
2. In all biota samples collected in Svalbard concentrations of Cd and Pb were below the limit of quantification of the used method. The average concentrations of Ni, Cu and Zn in samples of biota collected in Svalbard were comparable to the concentrations of nickel in samples collected in Poland.
3. The statistical assessment of the results and their visualization as a dendrogram show that the studied plant species accumulate heavy metals in a different ways, also in comparison to the surface soil layer. No statistical interrelations, which confirm the influence of local emission on heavy metals accumulation in the studied plants, have been found. Positive correlations in concentration pairs Ni/Zn accumulated in mosses, Mn/Cu accumulated in Thamnolia vermicularis and Mn/Zn and Cu/Zn accumulated in Salix Polaris were found.

We have executed experimental studies aiming to simulate the effects of various scenarios of winter disturbance. These studies were combined with simulation of air pollution, because northern vegetation is also exposed to air pollution from local or remote sources. The intention with adding air pollution to the experimental design is to simulate a realistic, dampened pollution load, which alone do not cause much harm, but may or may not accentuate the negative impacts of other stress, here winter perturbation. We have mostly used nitrogen as the pollution source. While nitrogen may stimulate growth, it can also reduce plant frost hardiness. As stated in the project’s description from 2012, the experimental studies was to be linked to other research projects financed by the Research Council of Norway and headed by Dr. Bjerke from NINA. This was done by using many of the same species that had been cultivated from seeds or fragments, and by using the same experimental designs. This included experiments in the field and in climate-controlled chambers. These studies were supplemented with additional freezing experiments in order to include a wider range of species. We measured mortality (alive-to-dead ratios), physiological responses (respiration, chlorophyll fluorescence, growth, electrolyte leakage, and optical measurements of nitrogen and chlorophyll content) and adaptive responses (gene expression and fatty acid changes).

Our experiments show that northern evergreen vascular plants experienced higher mortality than northern grasses and deciduous shrubs following full exposure to freezing winter air after a simulation of a winter warming event causing full snow disappearance. This was despite strong efforts of the evergreens to cope with the stressful situation; for example through a rapid and extensive expression of the CBF gene (C–repeat Binding Factor gene), which is involved in frost resistance. Ecophysiological responses to winter temperature stress varied greatly between and within plant functional groups and was not strongly correlated with plant mortality. Fatty acid composition of cell membranes changed with degree of winter disturbance for all grasses and deciduous shrubs, but only for one evergreen species. However, change was not in the same direction for all species. Some species, for example an arctic-alpine Festuca grass produced more of short-chained fatty acids, while an arctic-alpine Poa species produced more of both shorter and longer unsaturated fatty acids. The addition of simulated nitrogen pollution had only minor physiological effects in a few studied species and did not affect mortality.

Overall, these results reveal the vulnerability of northern evergreen plants to changing winter conditions. As such, evergreen plants may be at a great disadvantage, compared to other functional plant groups, when facing future winter climate conditions.

In an experiment using two High-Arctic and one temperate plant species we combined the stress from ice encasement and late frost, i.e. exposure to freezing conditions after a mild period in spring. With increasingly earlier onset of the growing season, the risk for freeze backlashes is increasing. Thus, the combination of winter disturbance and late frost stress is highly relevant. The two stresses were combined in series, meaning that surviving plants were first exposed to ice encasement and afterwards to late frost after a period of recovery.

While the temperate plant (a Polish variety of wild strawberry, Fragaria vesca) was very intolerant of ice encasement, the two high-Arctic species (the deciduous prostrate shrub Salix polaris and the graminoid Luzula confusa) showed no mortality on individual level; there were only some signs of top wilting. The few surviving shoots of wild strawberry died during the second stress exposure, i.e. the simulation of late frost. The two high-Arctic plants also experienced increased mortality after late frost. Eight per cent of L. confusa plants died completely, and surviving plant experienced a 31 % reduction of green biomass. The ratio of S. polaris plants with wilting top shoots increased by 58 percentage points to 81 % from the ice encasement to the late frost treatment. Total length of wilted shoots per plant was nearly 10 times higher after late frost than after ice encasement, and total leaf biomass was reduced by 49 %.

Although this selection of species only makes up a small fraction of the total number of temperate and Arctic species, it indicates that Arctic plants are more tolerant of winter disturbance than temperate plants. Results from other similar studies support this conclusion. Nevertheless, this study shows that High-Arctic plants are indeed affected by these types of disturbance, especially by the combined effect of winter icing and late frost in spring. Hence, earlier leaf-out in spring will increase the vulnerability also of Arctic plants, and this may affect future abundance ratios of species with variable tolerance to these stress types.

We have also tested the vulnerability of temperate, boreal and Arctic cryptogams to changing winter climate. A huge dataset involving gene expression, fatty acid composition, ecophysiology and survival from the same winter disturbance-air pollution experimental design as described above, is under preparation, and results are currently not ready to be summarized. However, we have other interesting results that are complete.

Using a field experimental design initiated already in 2007, we could study the longer-term impacts of winter warming events on a widespread Arctic-boreal moss. This moss was damaged by three consecutive winters with simulated warming events. We hypothesized that this moss would improve growth during the following years. However, even four years after the last winter warming event, photosynthesis and segment growth were still 30 and 36% lower than control levels, which was only a slight improvement from 44 and 43% 4 years earlier. Overall, these legacy effects demonstrate that this widespread and important moss is likely to be significantly disadvantaged in a future northern climate where frequent winter warming events may become the norm. Given the key importance of mosses for soil insulation, shelter and carbon sequestration in high-latitude regions, such persistent impacts may ultimately affect important ecosystem functions.

In a previous icing study (before this project), we showed that reindeer forage lichens were vulnerable to ice encasement. The lichens tested were continental species from areas where winters generally have a stable snow cover. In this project, to test if physiologically similar reindeer lichens from areas with more unstable snow cover were less vulnerable to winter climate perturbation, we selected an oceanic reindeer lichen, Cladonia portentosa, which grows on moors and heaths along to North Atlantic coast. This lichen also died during ice encasement, but its health decline was postponed by ca. 7 d compared to the continental species. This suggests that this species is slightly better adapted to anoxic conditions than its continental and more northerly counterparts, the lichens C. rangiferina and C. stellaris. It is likely that C. portentosa is more frequently encased in ice under natural conditions than species growing in more continental habitats.

We have also simulated winter disturbance in field experiments, involving several Arctic-alpine moss and lichen species. These studies varied in design, some have been with physical removal of snow and addition of liquid water to produce ice, and some have been with full snow thaw using infrared heaters. In these field studies, lichens and mosses are tolerant of ground-icing when the ice is covered by natural snow accumulation. The cryptogams, lichens in particular, are tolerant of full exposure to ambient air. Overall, quite extreme winter perturbation situations are required to completely kill lichens and mosses. Hence, we conclude that high-latitude cryptogams are more tolerant of winter climate change than the vascular plants with which they compete for space. Cryptogams on the ground may therefore benefit from increasing damage to vascular plants, despite some longer-term negative effects on moss growth, as those describe above. Arboreal lichens, on the other hand, are more vulnerable to winter perturbation, irrespective of origin. For example, the lichen Hypogymnia physodes, common on trees in temperate and boreal regions, quickly die when encased in ice. It has a special physiology, with air-filled internal cavities, which are important for gas circulation. When these are filled with ice, air circulation is hampered, and the green algae, which are active at near-zero subfreezing temperature, undergo anoxic respiration, which is lethal.

Some of the results presented here have been published or is under review, others are in manuscript form soon ready to be submitted to scientific journals. Although the project has formally ended, we will continue the processing of manuscripts from this work package.

Task 7.1. Analysis of spatial and temporal variability of air pollution over Polish and Svalbard test sites based  on archival ENVISAT SCHIAMACHY data, as well as on information acquired from ground stations in Poland

In order to analyze air pollution within forest study areas in Poland archival data concerning content of two pollutants – sulphur dioxide -SO2 and nitrogen dioxide – NO2 were collected from two ground stations located within Borecka Forest and Karkonosze for 2000 – 2014 period. These data were analyzed both in spatial and temporal aspect. Temporal analysis of SO2 and NO2 pollutants revealed their seasonal increase in late autumn – winter time, when analyzing annual changes. Comparison of yearly pollution levels from 2000 till 2014 did not reveal distinct trends, except NO2 in  Borecka   Forest, which level increased quite significantly in 2007 – 2014 period. In general both pollutants have higher values for Karkonosze study area than for Borecka Forest.

At the next stage of the works satellite data collected by ENVISAT SCHIAMACHY instrument were gathered for both study areas in Poland (only data related to NO2 content were available from 2003 – 2011 period). Analysis of these data confirmed conclusions drawn from analysis of ground observations – higher NO2 values for Karkonosze study area, no visible trend in the analyzed period and seasonal increase in wintertime. Correlation analysis of two datasets  - ground observations and satellite measurements revealed their moderate relationship (r=0.47).

Task 7.2. Analysis of relationships between climate variations (meteorological events) and spatial and temporal variability of vegetation indices derived from satellite images

Study of relationships between meteorological parameters and vegetation indices derived from NOAA AVHRR images  led to conclusion, that there is quite significant relation between these two types of data. Both indices – Normalized Difference Vegetation Index (NDVI) and Vegetation Condition Index (VCI) derived from first part of vegetation season correlate well with temperatures existing in wintertime, especially at the end of winter (in March). It means, that low-resolution satellite data can be applied for monitoring stress conditions at the beginning of growing season. Nevertheless, it should be mentioned, that significant impact of unfavourable conditions expressed by low winter / early spring temperatures is not observed for the study areas, while analyzing vegetation indices in the whole growing season. It implies the conclusion, that Polish forests located in both climatic zones are quite resistant to anomalies of temperature at the onset of vegetation season.

Similar study has been performed for vegetation indices derived from high-resolution Landsat satellite  data. At first thorough analysis of two years differing in meteorological parameters - 2014 and 2015 was carried out.  The dedicated hydrothermal index, based on temperature and precipitation data, which characterizes drought conditions, was calculated for all decades of growing season and compared with values of vegetation indices obtained from Landsat images. The analysis revealed, that in case of 2015 (characterized by long drought periods) decrease of DSWI (Disease water stress index) was much more higher (27 % for six species) than in case of 2014 (5 %). It points out, that vegetation indices derived from high-resolution satellite data can be an indicator of stress conditions in forests due to drought impact. This conclusion was supported by regression analysis between hydrothermal index from 2000 – 2016 period and decrease of DSWI index based on Landsat data, which revealed a good relationship between these variables (r = – 0.72). Moreover it was found, that the level of impact of drought conditions on forest is dependent on type of forest site; forests located on dry sites proved to be less susceptible to drought impact than those situated on fresh or humid sites.

Task 7.3. Analysis of relationships between vegetation indices and other metrics derived from satellite images and ground remote sensing measurements

Spectrometric measurements, conducted within ground campaigns in three growing seasons, were the basis for generating numerous vegetation indices, characterizing various aspects of plant condition and behaviour: general plant condition, chlorophyll content,  amount of dry material, plant senescence, amount of pigments, amount of nitrogen, amount of water. Among 40 indices six of them, which proved to change statistically significantly throughout the growing season and  correlated with vegetation parameters, have been selected: Normalized Difference Infrared Index (NDII) sensitive to water content in plants, Simple Ratio Index (SR) characterizing general plant condition, Enhanced Vegetation Index (EVI) sensitive to vegetation canopy structure, narrowband Normalized Difference Vegetation Index (NDVI 705) and Green Normalized Difference Vegetation Index (Green NDVI) characterizing chlorophyll content and  Simple Ratio Pigment Index (SRPI) sensitive to pigment concentration. The results of analysis of these indices in two contrasting seasons – 2014 and 2015 allowed to conclude, that some ground-based indices – NDII and SR reveal differences in changes of these parameters in the analyzed years, which can be related to climatic conditions – higher decrease of most indices appeared in 2015 (with exception of conifers and birch). Hence conclusions drawn from study of ground-based indices support results of analysis of high-resolution satellite images.

The study of relations between ground spectrometer (ASD FieldSpec) measurements and satellite data on Svalbard was aimed at assessment of different vegetation indices and their ability  to detect combined stress on vegetation. It was found an indication that combined impacts of pollution from the mines in Svalbard and winter warming events (in 2012) have had detrimental effects on the dwarf shrubs like the species Cassiope tetragona and Dryas octopetala.  A regression analysis of the in situ-derived RENDVI (ASD Fieldspec) of the species Dryas octopetala and the RapidEye-extracted RENDVI for the different sites dominated by this species showed a strong relationship of R2 = 0.82, p=0.036. This shows that it is possible to upscale from field-based assessments of vegetation stress (both pollution and climatic stress) to local and regional levels.

Task 7.4. Spatial analysis in north-south transect (Svalbard – Poland) of distribution of relationships identified in tasks 7.1 and 7.2 in order to determine impact of climatic changes on vegetation condition, depending on latitude and type of climate

Analysis of climate changes on vegetation conditions, performed at temperate climatic zone within forest study areas in Poland revealed that there is a relation between vegetation index derived for forests situated in northeastern Poland, characterized by continental climate and for those located in southern Poland, influenced partly by maritime climate. At the same time it was found that extreme changes of meteorological parameters, causing drought conditions can be reflected by decrease of dedicated vegetation index (DSWI).

However, it was found that Polish forests in a temperate zone are quite resistant to anomalies of temperature at the beginning of vegetation season, compensating unfavorable conditions  at the later stage of plant development.

In case of arctic climate on Svalbard it was found that dwarf vegetation is more vulnerable to the impact of climatic events and air pollution. Assessment of data from the Formosat satellite acquired in 2011 and 2012 reveals that the winter warming event (ground icing) in 2012, which caused detrimental damage to the dwarf shrub vegetation, can be detected on the 2012-image. Assessments of the high-resolution (Landsat 8) and low-resolution satellite data (MODIS, AVHRR) over the same sites (Adventdalen) for evaluating  vegetation health back in time show decreased maxNDVI (AVHRR) for the years 2005-2007, 2010 and 2012 which coincide with reported ground-icing episodes/winter warming events these years. So in general one can conclude, that impact of climate changes on vegetation condition is more pronounced in the northern latitudes than in a temperate climatic zone.

Task 7.5. Comparison of content of pollutants in plants in various study areas

The results of spatial analysis of content of pollutants in plants in various study areas were presented in the form of maps, where levels of particular heavy metals (nickel, zinc, cadmium, lead, manganese, copper) at measurements points were presented. For each study area the separate set of maps from succeeding field campaigns was prepared  and regions more pronounced to heavy metal accumulation were determined. Analysis of changes of metal concentration in plant elements collected during growing seasons showed no clear time trends, with the exception of Bialowieska Forest, where in 2015 growing season an increase of nickel concentration was observed. The most atmospheric aerosol pollution with heavy metal was found in Beskid Zywiecki Mountains (mainly with Zn, Cd and Pb), then in Karkonosze Mountains, the lowest levels of metal concentration were in northeastern forests (Borecka Forest).

Task 7.6. Examination of relationships between content of pollutants in plants and the condition of vegetation determined by ground, airborne and satellite remotely sensed measurements

Examination of relationships between contents of pollutants in plants and condition of vegetation, as described by ground measurements, was done through comparison of pollution concentration with levels of chlorophyll in plants.  Maps demonstrating spatial distribution of pollution  were the source of selection of sites for analysis. Distribution of pollution deposition was analyzed both in springtime and in summertime. The mean level of chlorophyll content for each species (spruce, beech, pine, hornbeam) was determined separately for the points with the highest concentration of pollutants and for the points with the lowest concentration of pollutants, in order to analyze, if there is any significant decrease in chlorophyll content in plants due to increased amount of pollutants. The results of this analysis reveal that there are no significant differences between chlorophyll level for plants affected with high pollution and those with the lowest concentration of pollution (with some exceptions in Karkonosze and Beskid Zywiecki Mountains, where slight decrease of chlorophyll content was observed in spring for spruce and beech).

Task 7.7. Analysis of degree of impact of air pollution on increased vulnerability of plants to climatic variations

Verification of hypothesis that vegetation areas affected by air pollution are more sensitive to climatic events like winter warming  was difficult within WICLAP project. For Polish test sites quite low level of pollution by heavy metals was observed, hence the impact on vegetation condition was in general negligible, as it was confirmed within task 7.6. In case of Svalbard test sites some relations between high contamination by nickel and intensity of damage of  tundra vegetation were found, but this relation will be further studied beyond WICLAP project, in order to make this hypothesis fully reliable.

Tømmervik H., Karlsen S.R., Nilsen L., Johansen B., Storvold R., Zmarz A., Beck P.S., Johansen K.S., Høgda K.A., Goetz S., Park T., Zagajewski B., Myneni R.B. & Bjerke J.W., 2014. Use of unmanned aircraft systems (UAS) in a multi-scale vegetation index study of Arctic plant communities in Adventdalen on Svalbard. EARSeL eProceedings, 13, S1: 47–52. doi: 10.12760/02-2014-1-09

Kłos A., Bochenek Z., Bjerke J.W., Zagajewski B., Ziołkowski D., Ziembik Z., Rajfur M., Dołhańczuk-Śródka A., Tømmervik H., Krems P., Jerz D. & Zielińska M., 2015. The use of mosses in biomonitoring of selected areas in Poland and Spitsbergen from 1975 to 2014. Ecological Chemistry and Engineering S 22: 201–218. doi: 10.1515/eces-2015-0011

Bochenek Z., Ziółkowski D., Bartold M., 2015. Application of NOAA AVHRR satellite images for studying various environmental and climatic conditions in Polish forests. Geoinformation Issues, Vol. 7 No 1(7), 29-37.

Bartold M., 2015. Cross-comparison of vegetation index time series from NOAA AVHRR and SPOT VEGETATION satellite observations to identify forest cover change in Poland. Sylwan 160 (2) pp. 153-161.

Bartold M., 2016. Development of forest cover mask to monitor the health condition of forests in Poland using long-term satellite observations. Lesne Prace Badawcze 77(2)  141-150. doi-10-1515-frp-2016-0016

Bokhorst S., Pedersen S.H., Brucker L., Anisimov O., Bjerke J.W., Brown R.D., Ehrich D., Essery R.L.H., Heilig A., Ingvander S., Johansson C., Johansson M., Jónsdóttir I.S., Niila I., Luojus K., Macelloni G., Mariash H., McLennan D., Rosqvist G.N., Sato A., Savela H., Schneebeli M., Sokolov A., Sokratov S.A., Terzago S., Vikhamar-Schuler D., Williamson S.N., Qiu Y. & Callaghan T.V., 2016. Changing Arctic snow cover: A review of recent developments and assessment of future needs for observations, modelling and impacts. Ambio 45: 516–537. doi: 10.1007/s13280-016-0770-0

Phoenix G.K. & Bjerke J.W., 2016. Arctic browning: events and trends as drivers. Global Change Biology 22: 2960-2962. doi: 10.1111/gcb.13261

Vikhamar-Schuler D., Isaksen K., Haugen J.E., Tømmervik H., Luks B., Schuler T.V. & Bjerke J.W., 2016. Changes in winter warming events in the Nordic Arctic Region. Journal of Climate 29: 6223–624. doi: 10.1175/JCLI-D-15-0763.1

Anderson H.B., Nilsen L., Tømmervik H., Karlsen S.R., Nagai S. & Cooper E.J., 2016. Using ordinary digital cameras in place of near-infrared sensors to derive vegetation indices for phenology studies of High Arctic vegetation. Remote Sensing 8: 847 (17 pp.). doi: 10.3390/rs8100847

Vickers H., Høgda K.A., Solbø S., Karlsen S.R., Tømmervik H., Aanes R. & Hansen B.B., 2016. Changes in greening in the high Arctic: insights from a 30 year AVHRR max NDVI dataset for Svalbard.  Environmental Research Letters 11: 105004.. doi: 10.1088/1748-9326/11/10/105004

Bjerke J.W., Treharne R., Vikhamar-Schuler D., Karlsen S.R., Ravolainen V., Bokhorst S., Phoenix G.K., Bochenek Z. & Tømmervik H. 2017. Understanding the drivers of extensive plant damage in boreal and Arctic ecosystems: Insights from field surveys in the aftermath of damage. Science of the Total Environment 599-600: 1965–1976.. doi: 10.1016/j.scitotenv.2017.05.050

Bochenek Z., Ziolkowski D., Barttold M., Orlowska K., Ochtyra A., 2016. high-resolution satellite data for monitoring forest areas in changeable climatic conditions.  European Journal of Remote Sensing (submitted).

Raczko E., Zagajewski B., 2017. Comparison of Support Vector Machine, Random Forest and neural network classifiers for tree species classification on airborne hyperspectral APEX images. European Journal of Remote Sensing, Vol. 50, Issue 1, 144-154.

Zagajewski B., Tømmervik H., Bjerke J.W., Bochenek Z., Orłowska K., Kłos A., Jarocińska A., Ziółkowski D., 2016. Intraspecific differences in hyperspectral reflectance curves as indicators of reduced vitality in High-Arctic plants. Remote Sensing (submitted).

Zagajewski B., Wietecha M., Bochenek Z., Kłos A., Bjerke J.W., Tømmervik H., Kycko M., Jarocińska A., Ochtyra A., Raczko E., 2017. Field measurements of spectral characteristics in health condition monitoring of forests in the Beskid Żywiecki region. Mountain Research and Development (submitted).

Orłowska K., Ochtyra A., Kycko M., Bochenek Z., Zagajewski B., Ziółkowski D. 2017. Hyperspectral detection of 2014 phenological changes in forest species of UNESCO's Bialowieza National Park. European Journal of Remote Sensing (submitted).

Kłos A., Ziembik Z., Rajfur M., Dołhańczuk-Śródka A., Bochenek Z., Bjerke J.W., Tømmervik H., Zagajewski B., Ziółkowski D., Jerz D., Zielińska M., Krems P. and Godyń P., 2016. The origin of heavy metals and radionuclides accumulated in the soil and biota samples collected in Svalbard, near Longyearbyen. Environmental Monitoring and Assessment (submitted).

Kłos A., Ziembik Z., Rajfur M., Dołhańczuk-Śródka A., Bochenek Z., Bjerke J.W., Tømmervik H., Zagajewski B., Ziółkowski D., Jerz D., Zielińska M., Krems P., Godyń P. Marciniak M. and Świsłowski P., 2017. Biomonitoring of contamination with heavy metals of forest areas in southern and north-eastern Poland. Chemosphere (submitted).

Kłos A., 2017. Monograph: Mchy w biomonitoringu środowiska (Mosses in environmental biomonitoring), Chapter 11: Badania biomonitoringowe realizowane w ramach projektu WICLAP (Biomonitoring studies carried out under the project WICLAP). WN PWN,  April 2017