Examensarbete vid Institutionen för geovetenskaper
Degree Project at the Department of Earth Sciences
ISSN 1650-6553 Nr 373
Variable Specific Discharge and Its Influence on Mass Export of Carbon, Sulphur, Calcium and Magnesium in a Boreal Forest Catchment
Variabel specifik avrinning och dess påverkan på exporten av kol, svavel, kalcium och magnesium från ett avrinningsområde i barrskogsbältet
Isabell Gärtner
INSTITUTIONEN FÖR GEOVETENSKAPER
D E P A R T M E N T O F E A R T H S C I E N C E S
Examensarbete vid Institutionen för geovetenskaper
Degree Project at the Department of Earth Sciences
ISSN 1650-6553 Nr 373
Variable Specific Discharge and Its Influence on Mass Export of Carbon, Sulphur, Calcium and Magnesium in a Boreal Forest Catchment
Variabel specifik avrinning och dess påverkan på exporten av kol, svavel, kalcium och magnesium från ett avrinningsområde i barrskogsbältet
Isabell Gärtner
ISSN 1650-6553
Copyright © Isabell Gärtner
Published at Department of Earth Sciences, Uppsala University (www.geo.uu.se), Uppsala, 2016
Abstract
Variable Specific Discharge and Its Influence on Mass Export of Carbon, Sulphur, Calcium and Magnesium in a Boreal Forest Catchment
Isabell Gärtner
Considerable research efforts are made in order to understand the global carbon cycle and how it will affect future climate change and vice versa. To be able to calculate the export of carbon from a certain area, discharge is one of the most important variables together with stream concentrations. Measuring discharge in every catchment would be impossible, as it is both time consuming and expensive. To come around these obstacles, the majority of studies on element export use known discharge data from gauging stations at a single catchment outlet and assumes the same discharge per unit area from nearby catchments, known as the assumption of uniform specific discharge. A few studies in recent years, have come to the conclusion that this questionable assumption can lead to large errors in estimated discharge volumes and it should therefore be reconsidered.
This study aims to analyse how the export of elements changes if actual measured variable discharge is applied in comparison to uniform specific discharge. The results of this study shows that the assumption of uniform discharge leads to an overestimation of the export of dissolved organic carbon (DOC), sulphur, calcium and magnesium from forest-dominated catchments by up to 30%. At the same time this assumption leads to an underestimation of export from wetland-dominated catchments by up to 26% over a five year period.
Keywords: Hydrochemistry, hydrology, specific discharge, boreal catchment Degree Project E1 in Earth Science, 1GV025, 30 credits
Supervisor: Thomas Grabs
Department of Earth Sciences, Uppsala University, Villavägen 16, SE-752 36 Uppsala (www.geo.uu.se) ISSN 1650-6553, Examensarbete vid Institutionen för geovetenskaper, No. 373, 2016
The whole document is available at www.diva-portal.org
Populärvetenskaplig sammanfattning
Variabel specifik avrinning och dess påverkan på exporten av kol, svavel, kalcium och magnesium från ett avrinningsområde i barrskogsbältet
Isabell Gärtner
Mycket av forskningen de senaste åren har handlat om framtida klimatförändringar och det globala kolkretsloppet och hur de båda påverkar varandra. För att göra tillförlitliga beräkningar av hur mycket kol ett område, i det här fallet ett barrskogsområde i norra Sverige, tillför vattendragen i närheten under en viss tid skulle det behöva göras vattenståndsmätningar och vattenanalyser vid varje vattendrag, men det skulle vara omöjligt eftersom det är för kostsamt och tidskrävande. För att ändå kunna göra uppskattningar av hur mycket kol, spårämnen och metaller som kommer från ett område, använder de flesta studier information om vattenstånd och ämneskoncentrationen i vattnet från mätningsstationer vid avrinningsområdets utlopp. Samma uppgifter används för att göra beräkningar för det stora området, som för mindre delområden i det. Under de senaste åren har ett antal studier kommit fram till att användandet av data som inte har tagits i direkt anslutning till ett avrinningsområde kan leda till stora fel och kanske borde undvikas.
Den här uppsatsen har som syfte att analysera vilka skillnader som uppstår när avrinningsdata från olika delområden används istället för samma data för alla områden. Resultatet av denna studie visar att om man använder samma data för alla områden leder det till att exporten av kol, svavel, kalcium och magnesium från huvudsakligen skogsklädda områden överskattas med upp till 30% och underskattas från områden med mycket våtmarker med upp till 26%.
Nyckelord: Hydrokemi, hydrologi, avrinning, avrinningsområde i barrskog Examensarbete E1 i geovetenskap, 1GV025, 30 hp
Handledare: Thomas Grabs
Institutionen för geovetenskaper, Uppsala universitet, Villavägen 16, 752 36 Uppsala (www.geo.uu.se) ISSN 1650-6553, Examensarbete vid Institutionen för geovetenskaper, Nr 373, 2016
Hela publikationen finns tillgänglig på www.diva-portal.org
Table of Contents
1. Introduction ... 1
2. Aim ... 2
3. Background ... 3
3.1 Previous studies ... 3
3.1.1 Processes affecting the export of dissolved organic carbon (DOC) in boreal catchments ... 3
3.1.2 Processes affecting the export of sulphur, calcium and magnesium in boreal catchments .... 4
3.1.3 Runoff generation processes and the assumption of uniform specific discharge ... 5
4. Methodology... 7
4.1 Study site ... 7
4.1.1 Introduction ... 7
4.1.2 Soil, geology and vegetation ... 8
4.1.3 Climate ... 9
4.2 Calculation of mass export rates... 10
4.3 Analyses of discharge concentration relationships ... 11
5. Results ... 11
5.1 Whole time series and annual results ... 11
5.2 Seasonal and monthly results ... 16
5.3 Result of discharge-concentration relationship ... 20
6. Discussion ... 24
7. Conclusion ... 26
8. Acknowledgments ... 27
9. References ... 28
Appendix ... 30
1
1. Introduction
Research on climate change in the last decades has been focusing on the importance of oceans and the atmosphere and their role in the carbon cycle. In recent years the boreal landscape has come more into focus. Not only does it cover vast areas in the higher latitudes, it also stores about 30% of the terrestrial carbon (Laudon et al. 2013) . This means that the fate of the boreal landscape plays not only an important role on a local but on a global scale as well, affecting significantly the carbon cycle and element release to the oceans.
Future climate change is expected to cause a wetter climate in the Northern Hemisphere, with shorter snow season, larger winter base-flow and less intense spring flood (Teutschbein et al. 2015).
The impact of these changes might differ from catchment to catchment, but also on the subcatchment scale. A change of the climatic conditions could have an impacts on hydrological and hydrochemical processes on smaller and larger scales that are difficult to predict (Laudon et al. 2013)
Hydrological processes are not limited to how much discharge is generated in a particular catchment as a response to a certain amount of precipitation, but it affects many processes within those catchments. Precipitation and runoff generation also influence the export of elements, dissolved organic carbon (DOC) and sediments, with impacts on biological activity (Teutschbein et al. 2015). In order to make predictions of how for example element export will change in the future it is important to have a solid understanding of how these processes work today, what influences them and which feedback mechanism comes into play. Discharge is a key variable for calculating element export from catchments and their subcatchments. In practice it is common to use discharge data from only one point of the catchment, for example the outlet. This is because data from smaller streams is usually missing and would be far too costly to collect and manage. To still be able to estimate discharge from subcatchments, one approach is to use the assumption of uniform specific discharge, which means that the same discharge data per unit area is used for element export calculation for all subcatchments.
Using uniform specific discharge instead of spatially variable specific discharge is a simplification that was considered to be valid and the effect on the element export only minor. Catchments with similar physical properties, for example landscape type, soils, vegetation, topography amongst others, were assumed have similar discharge and consequently similar export of trace elements and DOC.
Although research in recent years (Karlsen et al. 2016; Teutschbein et al. 2015; Lyon et al. 2012) has indicated that this might not be the case and that even small differences between otherwise similar subcatchments might have significant impact on runoff. In cases where both element concentration and discharge data from subcatchments are available, reliable mass export rates can be calculated to analyse the consequences of using the assumption of uniform specific discharge. If large differences in specific discharge were found, the results of previous studies regarding hydrology, hydrochemistry (Lidman et al. 2014; Mörth et al. 2008) and carbon cycle (Ågren et al. 2007; Ågren et al. 2014;
Björkvald et al. 2008; Laudon et al. 2011; Ledesma et al. 2015) must be questioned.
2
2. Aim
This study aims to investigate the potential differences in element export that arise from calculations using uniform specific discharge and spatially variable specific discharge. The questions this study tries to answer are:
• Does the element export per unit area differ when variable discharge is applied in comparison to uniform discharge, and if so by how much?
• Are these possible differences connected to physical properties of the catchment, like landscape type?
• Do differences occur year round or are they bound to a certain season or hydrological processes, like the spring flood?
To address these questions, discharge and chemical data from 14 nested subcatchments of the Krycklan catchment, Northern Sweden, were analysed over a period of five years. These subcatchments have similar physical properties and in the vast majority of previous studies, uniform specific discharge was assumed. Although Krycklan is a comparatively small catchment, the same assumptions are widely used even for larger catchments and other landscape types. The elements chosen for the element export calculations are carbon (C), sulphur (S), calcium (Ca) and magnesium (Mg).
3
3. Background
The literature study focused mainly on articles published in recent years, that included the Krycklan catchment or that used data from the Krycklan catchment study. Of the large number of published articles a limited number of articles that focus on DOC, sulphur and metal export as well as specific discharge were chosen.
3.1 Previous studies
3.1.1 Processes affecting the export of dissolved organic carbon (DOC) in boreal catchments
A large number of studies have investigated if certain landscape features are connected with higher or lower DOC leaching (Ågren et al. 2014), possible carbon pools and their turnover time (Ledesma et al.
2015), the connectivity of the hydrological system (Laudon et al. 2011), the DOC concentration of headwaters (Ågren et al. 2007) and larger streams and in-stream processes like photosynthesis that could degrade organic matter (Tiwari et al. 2014). Dissolved organic carbon (DOC) is a degradation product from both animals and plants. About 90% of the carbon in lakes and rivers occurs in either completely dissolved form or small clusters or colloids, between 0.2 and 0.45 µm in size. Very small colloids like these will stay suspended in the water and not sink to the bottom and sediment even in calm waters (Tranvik and von Wachenfeldt 2009). DOC can together with iron form soluble colloids or organic complexes and by that lowering the concentrations of these solutes in the stream water (Björkvald et al. 2008).
The concentration of DOC in the stream is dependent on a number of factors. The single most important one seems to be if the catchment is wetland- or forest-dominated. Wetland-dominated catchments have generally a higher export of DOC (Ågren et al. 2007), but also the highest spatial variability between 2-41 mg/L (Laudon et al. 2011). The variability is clearly linked to the different seasons in the area, the export of DOC is low during winter base-flow, but higher than in forest- dominated catchments and increases in the spring, but less than for other catchment types. One plausible explanation could be that during winter the anoxic environment in the peat provides a source of carbon, and with the onset of snow melt the stream water is diluted (Björkvald et al. 2008) as melt water enters the stream as overland flow without infiltrating the still frozen ground (Laudon et al., 2007). DOC in the stream increases also after rainfalls, probably because DOC stored in the peat soil is mobilized and transported to the stream during such events (Ågren et al. 2014).
During the spring flood the DOC export is similar for all catchments, which means that forest- dominated catchments export about 210% more DOC and wetlands 60% less (Laudon et al. 2011).
After the spring flood, at the beginning of the summer the DOC concentration in the stream increases again (Björkvald et al. 2008) , likely because of more biological activity in the catchment. In forested catchments about 90% of the DOC export originates from a relatively shallow, organic matter rich
4
layer, the dominant source layer (DSL) within the near-stream (riparian) zone. The riparian zone is usually larger in waterlogged soils underlain by till, with shallow groundwater tables (Ledesma et al.
2015). Silty catchments have the lowest export of DOC, which might be because the DOC is sorbed to mineral surfaces instead of transported further. Studies have also found indications that DOC export is related to the type of trees that are dominant in a certain catchment, therefore differences can occur even between two forested catchments (Ågren et al. 2014) .
Small headwaters at higher altitude in the Krycklan catchment seem to be the largest source of DOC per unit area (Ågren et al. 2007). DOC export is generally low during base-flow and higher during larger discharge, which means that there is an almost linear relationship between discharge and DOC export. The total amount of DOC that is exported is higher during periods of high discharge, even if the concentration in the stream might be lower (Ågren et al. 2007; Ågren et al. 2014; Ledesma et al.
2015). At the catchment outlet the DOC concentration is low during winter base-flow (Björkvald et al.
2008), this might be due to deep, DOC-poor groundwater flows (Tiwari et al. 2014), but reaches a peak at the start of the summer season (Björkvald et al. 2008).
The pathways of the water in the soil change with the wetness state of the system and by that regulate the amount of DOC that is exported. The connectivity decreases when the system becomes drier, and vice versa. Changes in flow paths due to a wetter climate could increase DOC export by activating carbon rich soil layers, but could also decrease in the long term when carbon pools are depleted. However, this seems unlikely as the DSL layers in the riparian zone alone contain enough carbon to sustain current DOC exports for several hundreds to thousands of years (Ledesma et al.
2015). The loss of DOC in the stream from the headwaters to the outlet, due to degradation and photosynthesis in the Krycklan catchment seems to be negligible (Tiwari et al. 2014). How future climate change would affect the DOC export is difficult to predict as some processes could lead to a positive feedback, while others to a negative one. Predictions for heterogeneous catchments are especially difficult to make (Laudon et al. 2011).
The export of DOC can vary substantially in boreal catchments. Export rates of 14 800 - 99 100 kg/km²a have been reported from the Krycklan catchment, under the assumption of uniform specific discharge (Ågren et al. 2007). Other boreal catchments might have very different export rates, for example a recent study in Southern Finland, with average export rates of 6100 kg/km² a (Huotari et al.
2013). Element export varies also substantially with seasons. In subarctic and arctic landscapes DOC export can increase 3 to 25 times during the spring flood. During the peak spring flood Alaskan arctic rivers can export more than 30% of the annually exported elements in less than two weeks (Rember and Trefry 2004).
3.1.2 Processes affecting the export of sulphur, calcium and magnesium in boreal catchments Elements, like calcium and magnesium, originate either from weathering of mineral soils and bedrock or deposition. Weathering is a process mainly restricted to forest floors. Elements are leached out of
5
the soil or the bedrock and then transported with the percolating water into deeper soil layers or directly to the streams. Weathering in wetlands is very limited. This kind of landscape on the other hand has shown the ability to bind and accumulate mainly organophilic metals, like uranium. Non- organophilic metals like calcium and magnesium are also affected (Lidman et al. 2014).
Organophilic metals bind, as the name implies, preferably to organic matter, which is more abundant in the peat soils of wetlands than the mineral soils of forest. Although magnesium and calcium are non organophilic, the concentration of calcium is decreasing by 1,1% and magnesium by 1,3% per percent wetland area of a catchment. A total of 20-40% of all elements is intercepted in wetlands on their way to the stream. Wetland-dominated catchments have therefore generally lower fluxes of elements than forest-dominated. Usually the effect is more pronounced in smaller catchments, but water fluxes and local hydrology are important as well. In simple terms, more runoff from forests via wetlands equals more bound elements and lower concentration in the stream (Lidman et al. 2014). In a study of element transport in the Krycklan catchment. Lidman et al. (2014) found the export rate of Ca and Mg to be 180-750 kg/km²a and 67-260 kg/km²a respectively, but these export rates can vary from catchment to catchment. Pogkrovsky et al. (2006) reported much higher export rates from a Siberian catchment of 3000 and 540 kg/km² a for Ca and Mg respectively.
The main source of sulphur in the stream is the weathering of sedimentary sulphides, erosion of acid sulphuric soils and anthropogenic deposition. The sources vary depending on the season, during base- flow the bacterial reduction of peat in wetlands and riparian zones are the main source of sulphur. At the start of the snow melt season, anthropogenic sulphur from the snow is added, but at the same time the melt water dilutes stream concentrations. Snow sulphur concentrations are usually low in the Northern part of Sweden. About 11-29 kg/km² a of sulphur originates in snow deposition of a total anthropogenic deposition of 140-180 kg/km² a. Sulphur chemistry in the soil and the stream is dependent on redox reactions, which can release or bind sulphur. Anoxic conditions in combination with organic matter in wetlands can for example act as sink for sulphur. At large scales, land use change and agricultural activity can influence the export of sulphur from a certain area (Mörth et al.
2008). At small scales Kortesky et al. (2007) showed that element concentrations in soil pore water differed markedly within 10 m² of wetland, most likely due to different kinds of vegetation affecting redox reactions and to the local biogeochemistry. Previous studies reported an S export rate of 390 kg/km² a for parts of the Krycklan catchment and 180-350 kg/km² a for other streams in Northern Sweden (Mörth et al. 2008).
3.1.3 Runoff generation processes and the assumption of uniform specific discharge
Discharge is a key element when estimating export rates, but monitoring small streams is expensive.
Therefore, data from the catchment outlet is used to make calculations for subcatchments, under the assumption that the discharge per unit area is the same throughout the entire catchments. Several studies, however found this assumption of uniform specific discharge to be questionable even in the
6
relatively homogenous region of the Krycklan catchment. Lyon et al. (2012) showed that during three separate sampling campaigns in the Krycklan catchment the specific discharge among subcatchments in Krycklan varied between 37-43 % from the median. Small catchments and especially the headwaters showed the highest variability. Karlsen et al. (2016) came to similar conclusions in a study of specific discharge of 14 subcatchments in the same area, with a five year time series of precipitation and runoff The discharge variability of the subcatchments in comparison to the catchment outlet was between 33% (daily scale) and 19% (annual scale).
Differences between variable and uniform specific discharge show the same general pattern (Teutschbein et al. 2015b) and have similar seasonal trends and might, therefore, not be solely explained by measuring errors or uncertainty (Karlsen et al. 2016). Variations in discharge of 20-30%
would otherwise be expected after large scale land use changes, for example deforestation, or the expected climate change at the end of the century. The largest variation in discharge appeared in smaller catchments and during periods of low flow, which might be due to differences in the wetness state of the system and evaporation (Karlsen et al. 2016).
The seasonal difference could be explained by a number of processes in the catchment affecting each other. Forests seem to evaporate large amount of soil moisture, making the soil in these parts of the catchment drier, whereas wetlands have the capacity to store moisture (Lyon et al. 2012). Leading to a different flow pattern and changes of the connectivity in the system (Laudon et al., 2011).
Variations in specific discharge can therefore occur not only between catchments, but also between wetter and drier months in the same catchment (Lyon et al. 2012).
Runoff generation is also dependent on landscape properties. Forest-dominated catchments contribute only about 10-30% of new or event water to the stream, whereas wetland-dominated catchments contribute 50%. The pathways water takes in the system seems to be different depending on the landscape type, for example frozen soil is more permeable in forests than in wetlands which allows more infiltration of melt water. In wetlands pooling of water and overland flow are common during the snow melt season, mainly because the ground is frozen solid (Laudon et al. 2011).
The predicted increase in temperature and precipitation are expected to lead to an increase in winter base flow, shorter winter season, less snow cover, and shorter time with snow covered ground and due to that less intense spring floods (Kjellström et.al 2014 ; SMHI 2015; Teutschbein et al. 2015).
Generally, both high and low flow during the year would decrease with fewer extreme flows. Forest- dominated catchments are expected to react differently to climate changes, for example having higher annual mean flow than wetlands. Wetland-dominated catchments are expected to have a larger decrease in the peak spring flood. A change in stream flow would having a multitude of impacts on hydrochemistry, ecology, water quality, but also on physical processes like erosion and export of sediments, carbon and other substances (Teutschbein et al. 2015).
7
4. Methodology
4.1 Study site
4.1.1 Introduction
The boreal Krycklan catchment is a well monitored study area, mainly for biogeochemical research, run by the Swedish University of Agricultural Sciences (SLU). The study site is located close to the city of Vindeln, about 50 km northwest of Umeå (Laudon et al. 2013). Krycklan is a 68 km² large catchment, with currently 18 monitored nested subcatchments. Data of 13 of these subcatchments, ranging in size from 0.12-19,13 km², as well as the 67.9 km² large Krycklan catchment that includes the smaller subcatchments were used in this study (Figure 1).
The first research efforts at Krycklan were made in the early 1900s, focusing at that time mainly on forestry and soil frost. In 1980 the Svartberget field station, at the outlet of C7, was established in Krycklan (Figure 1), generating data for hydrological and hydrochemical research (Laudon et al. 2013) The initial study area was extended over the years, until 2002, to its current size (Laudon et al. 2011).
Figure 1. Map of the Krycklan study site (C1, C2, C4-C7, C9, C10, C12-C16 and C20) with stream gauging stations and corresponding subcatchments (Laudon et al. 2007, map reproduced with permission from Ida Taberman, SLU Umeå).
N
C4 C10 C5 C2
C6 C7 C1 C15
C9 C12 C14 C13
C20
C16 Name of the
subcatchments:
C1: Risbäcken C2: Västrabäcken C4: Kallkälsmyren C5: Stortjärnen outlet C6: Stortjärnsbäcken C7: Kallkälsbäcken C9: Nyängesbäcken C10: Stormyrbäcken C12: Nymyrbäcken C13: Långbäcken C14: Åhedbäcken C15: Övre Krycklan C16: Krycklan C20: -
8 4.1.2 Soil, geology and vegetation
Soils in the more elevated parts of the catchment are mainly unsorted glacial till and peat (Laudon et al. 2011). The majority of the headwaters (small watercourses with stream order 1) (Nadeau and Rains 2007) are situated in the more elevated parts above the highest coast line. Below the highest coast line the soils are finer sandy and silty soils (Laudon et al. 2011). The catchment is dominated by forest (87%) and wetlands (9%). Lakes and cultivated land cover 1% of the study area each (Laudon et al.
2013). Detailed landscape properties are listed in table 1. The subcatchments are divided into forest- dominated (<2% wetland coverage), wetland-dominated (>30% wetland coverage) and mixed (2-30%
wetland coverage) according to previous studies in the area (Buffam et al. 2007; Björkvald et al.
2008).
The geology is mainly characterized by metasediments/metagreywacke (94%) and smaller areas of acid, intermediate and basic metavolcanic rocks (6%) (Laudon et al. 2013). Forest vegetation is dominated by Norway spruce (Picea abies) and Scots pine (Pinus sylvestris), as well as alder (Alnus glutinosa) and birch trees (Betula pendula) (Laudon et al., 2011), the forest floor is covered by shrubs and moss (Forsum et al. 2008), while wetlands and riparian zones are mainly vegetated by Sphagnum species (Ågren et al. 2014).
Table 1. Catchment properties (Laudon et al. 2007; Laudon et al. 2013).
Catchment Area (km²)
Stream order
Lakes (%)
Forest (%)
Open land (%)
Arable land (%)
Mire (%)
Till (%)
Thin soils (%)
Rock out- crops (%)
Sorted sediments (%)
C1 0.48 1 - 98.0 - - 2.0 92.1 7.9 - -
C2 0.12 1 - 99.9 - - - 84.2 15.7 - -
C4 0.18 1 - 55.9 - - 44.1 22.0 27.0 - -
C5 0.65 1 6.4 54.0 - - 39.5 40.4 5.5 - -
C6 1.10 1 3.8 71.4 - - 24.8 53.7 11.3 2.5 -
C7 0.47 2 - 82.0 - - 18.0 65.2 15.4 - -
C9 2.88 2 1.5 84.4 - - 14.1 69.1 6.8 1.7 4.1
C10 3.36 3 - 73.8 - - 26.1 59.9 10.8 - 0.5
C12 5.44 3 - 82.6 - - 17.3 66.6 8.4 - 5.9
C13 7.00 3 0.7 88.2 0.2 0.6 10.3 60.9 8.9 1.3 15.9
C14 14.10 3 0.7 90.1 0.9 2.9 5.4 44.9 8.1 1.6 38.1
C15 19.13 4 2.4 81.6 1.4 0.1 14.5 64.8 8.1 0.7 9.5
C16 67.90 4 1.0 87.2 1.1 1.9 8.7 50.8 7.4 1.2 30.2
C20 1.45 ? - 87.7 2.6 9.6 45.0 20.3 1.8 21.4
9 4.1.3 Climate
The Krycklan catchment is according to the Köppen-Geiger classification described as Dfc, a fully humid, snow climate with cold summers and cold winters like most of Scandinavia (Kottek et al.
2006). The annual mean temperature is +1.7°C and the annual mean precipitation 612 mm (1980- 2008), the annual mean runoff is 312 mm. About half of the precipitation is falling as snow and the ground is snow covered for at least five months every year from the end of October to the beginning of May (Laudon et al. 2011). Snow melt usually starts in mid-April, lasting for about 1.5 months (Teutschbein et al. 2015b).
The precipitation in the area varies seasonally with less snow and rainfall during late winter and early spring (March and April) and more rain during the summer months (June, July and August) (Figure 2), but the seasonal variation is less than for discharge (Figure 3). The total annual precipitation during the studied time series (October 2008 - September 2013) varied between the wettest year, year 4 (October 2011- September 2012), with 761 mm to the driest year, year 3 (October 2010 - September 2011), with an annual precipitation of 581 mm.
The two months with the highest discharge (April and May) are at the same time those with the least precipitation, whereas the summer and autumn months (July, August and September) are, meteorologically, the wettest. Heavy rainfall events are most likely responsible for the discharge peaks during these months (Figure 3).
Figure 2. Precipitation at Krycklan.
10 Figure 3. Discharge at the catchment outlet (C16).
4.2 Calculation of mass export rates
All data was provided by the Swedish Agricultural University (SLU), as part of the Krycklan catchment study (SLU, 2016). The data set included daily precipitation and runoff, as well as hydro chemical analysis of 13 subcatchments (C1-7, C9-15 and 20), and of the catchment outlet, C16, over a period of five years (October 2008- September 2013) . Sampling for the hydro chemical analysis was done monthly during the winter period, due to the difficult climatic conditions. In the autumn and summer the sampling was done about twice a month and during the spring twice a week. To obtain daily hydrochemical data, missing values were gap filled using linear interpolation. The obtained discharge data was first converted to units of liters per second, and then the flux (exper rates) of DOC, Ca, Mg and S for each catchment was calculated using equation (1). Five-year, annual, seasonal and monthly averages were then calculated for every element and catchment.
𝑄𝑄 �𝑚𝑚𝑚𝑚𝑑𝑑 � ∗ 𝐴𝐴 (𝑘𝑘𝑚𝑚2) ∗1000106 = 𝑄𝑄 �𝐿𝐿𝑑𝑑�/ 86400 = 𝑄𝑄 �𝐿𝐿𝑠𝑠� (1a) 𝑄𝑄 �𝐿𝐿𝑠𝑠� ∗ 𝐶𝐶 �𝑚𝑚𝑚𝑚𝐿𝐿 � = 1000𝑚𝑚𝑚𝑚𝑠𝑠 ∗ 86400 = �𝑚𝑚𝑑𝑑� ∗1000365𝐴𝐴 (𝑘𝑘𝑚𝑚2) = 𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸 𝐸𝐸𝑟𝑟𝐸𝐸𝑟𝑟 �𝑘𝑘𝑚𝑚𝑘𝑘𝑚𝑚2𝑎𝑎� (1b)
In the above equations, Q is discharge, and A is Area.
Throughout the thesis the term "uniform discharge" is used when discharge from one subcatchment outlet (C7) is used to make element export calculations for all subcatchments and the term "variable discharge" when data from one catchment outlets is used only to make calculations for this particular subcatchment. To discern possible differences between variable and uniform discharge, all calculations were done using the discharge of the individual catchment (variable discharge) and the
11
discharge of the Kallkälsbäcken or Svartberget subcatchment, C7, (uniform discharge). The differences between the variable and uniform export rates were then calculated as:
Difference in element export (%) =𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸 𝐸𝐸𝑎𝑎𝐸𝐸𝑟𝑟𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣 𝑄𝑄 ∗100
𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸 𝐸𝐸𝑎𝑎𝐸𝐸𝑟𝑟𝑢𝑢𝑢𝑢𝑣𝑣𝑢𝑢𝑢𝑢𝑣𝑣𝑚𝑚 𝑄𝑄 − 100 (2)
The Svartberget subcatchment (C7) was chosen over the entire Krycklan catchment (C 16), because this catchment has the longest, continuous, record of discharge measurements and the results can be more easily compared to previous studies. The seasonal division of the year was done according to the common practice for this region, i.e. autumn (September, October), winter (November, December, January, February, March), spring (April, May) and summer (June, July, August) (Wastenson et al.
1995). The time series of catchment C13 is slightly shorter (August 2008-September 3013) because of missing discharge data. Therefore, only data from 4 years (October 2009-September 2013) were used in the calculations from this catchment.
4.3 Analyses of discharge concentration relationships
To show the relationship between discharge and concentration, the observed values of both parameters were plotted against each other in a scatter plot. A power law function was fitted between variable discharge and concentration measured at the subcatchment outlets:
C = a𝑄𝑄𝑏𝑏 (3)
The exponent parameter b can be interpreted as showing the sensitivity of element concentration to changes in discharge, for example a negative parameter b indicates that the element concentration was diluted due to an increased discharge. In the mineral soil of the forest negative b values could mean deeper flow paths and more weathering. Wetter conditions with positive b values show likely shallower flow paths with more DOC leaching and less weathering. If b is equal to 0 the element concentration is independent from discharge (Seibert et al. 2009).
5. Results
5.1 Whole time series and annual results
The results of the relative differences in DOC export, between uniform and variable specific discharge (Figure 4) shows both positive and negative values, In wetland-dominated catchments (C4 and C5) the uniform specific discharge underestimated the DOC export by 14-26% and in forest-dominated
12
catchments (C1 and C2) overestimates the export by 9-30%. In the silt underlain catchment (C14) the export is overestimated by 25% and in the mixed catchments the export is both under- and overestimated.
Figure 4. Differences in DOC export for the whole five years time series, if variable discharge is compared to uniform discharge, calculated from the Svartberget catchment (C7). The catchments are sorted according to how much of the catchment is covered by wetland, with highest areal coverage to the left and lowest to the right.
Positive numbers in the diagram mean that using the assumption of uniform discharge the export is underestimated in this catchment and negative numbers mean the opposite.
The relative differences in element export for S, Ca and Mg (Figure 5) are very similar to the export rates of DOC. Export rates for wetland-dominated catchments are underestimated using uniform specific discharge and in forest-dominated catchments the export is overestimated.
Figure 5. Differences in S, Ca and Mg export for the whole five years time series, if variable discharge is compared to uniform discharge, calculated from the Svartberget catchment (C7). The catchments are sorted according to how much of the catchment is covered by wetland, with highest areal coverage to the left and lowest to the right. Positive numbers in the diagram mean that using the assumption of uniform discharge the export is underestimated in this catchment and negative numbers mean the opposite.
-40 -30 -20 -10 0 10 20 30
4 5 10 6 7 12 15 9 13 20 16 14 1 2
Differences in export (%)
Catchment
DOC
-40 -30 -20 -10 0 10 20 30
4 5 10 6 7 12 15 9 13 20 16 14 1 2
Differences in export (%)
Catchment
S Ca Mg
13
The differences in flow (Figure 6) show the same pattern as the element export of DOC, S, Ca and Mg (Figure 4 and 5) with only minor deviations in some catchments.
Figure 6. Differences in flow for the whole five years time series, if variable discharge is compared to uniform discharge, calculated from the Svartberget catchment (C7). The catchments are sorted according to how much of the catchment is covered by wetland, with highest areal coverage to the left and lowest to the right. Positive numbers in the diagram mean that using the assumption of uniform discharge the export is underestimated in this catchment and negative numbers mean the opposite.
The average export rate for DOC shows that wetland-dominated catchments (C4 and C5) have overall a higher export than other catchments and forest-dominated catchment have higher or similar export than mixed catchments (Figure 7).
Figure 7. Average export rates of DOC for the five years time series, using variable discharge. The catchments are sorted according to how much of the catchment is covered by wetland, with highest areal coverage to the left and lowest to the right.
-30 -20 -10 0 10 20 30
4 5 10 6 7 12 15 9 13 20 16 14
Flow (%)
Catchment
0 2000 4000 6000 8000 10000 12000
4 5 10 6 7 12 15 9 13 20 16 14 1 2
Export rate (kg/km²a)
Catchment
DOC
14
The five year average export rates of S, Ca and Mg (Figure 8) show slightly lower export for the wetland-dominated catchments, C4 and C5, (except Ca). The mixed catchment C20 had the highest export for all elements and years.
Figure 8. Average export rates of S, Ca and Mg for the five years time series, using variable discharge. The catchments are sorted according to how much of the catchment is covered by wetland, with highest areal coverage to the left and lowest to the right.
The export of DOC is generally higher from wetland-dominated sites (Figure 9a) , for both the variable and the uniform discharge, than from forest-dominated catchments. In the wetland-dominated sites the export of DOC is higher using the variable discharge then when assuming uniform specific discharge, whereas the opposite is the case in the forest-dominated catchments (Figure 9b). The wettest year of the time series (year 4) resulted in the highest export of DOC of the time series and the driest year (year 3) in the lowest. This pattern occurred for both catchment types and independent if variable or uniform discharge is used.
0 200 400 600 800 1000 1200 1400 1600 1800 2000
4 5 10 6 7 12 15 9 13 20 16 14 1 2
Export rates (kg/km²a)
Catchment
S Ca Mg
15
Figure 9. a) Average DOC export, wetland catchment (C4). b) Average DOC export, forest catchment (C2).
The annual export of S, Ca and Mg show a similar pattern to the DOC export in the forest-dominated catchment (Figure 10b), but for the wetland-dominated catchment (Figure 10a) the differences are only marginal, except for year 3. Higher or lower annual precipitation has lesser effect on the export rates, than it does for DOC.
a)
b)
16
Figure 10. a) Average Mg export, wetland catchment (C4). b) Average Mg export, forest catchment (C2).
5.2 Seasonal and monthly results
The export of DOC showed a clear seasonal pattern (Figure 11). The wetland-dominated catchment (C4) had the highest export during all seasons, whereas the forest-dominated catchment (C2), the silt underlain catchment (C14) and the catchment outlet (C16) show similar export rates. From winter to spring all catchment types had increased export rates. The forest-dominated catchment had the highest increase by 960 %, the wetland-dominated catchment the lowest, 310 %.
a)
b)
17
Figure 11. Changes in DOC export with season, using variable discharge. Season 1 = autumn (SO); Season 2 = winter (NDJFM); Season 3 = spring (AM); Season 4 = summer (JJA).
The wetland-dominated catchment (C4) and the forest-dominated catchment (C2) have the lowest export of S, Ca and Mg during all season (Figure 12). The export is higher during autumn and especially during the spring flood. The silt underlain catchment (C14) and the catchment outlet (C16), show all a similar pattern with medium high export rates during autumn and summer and a strong increase during the spring from very low values during the winter.
Figure 12. Changes in Ca export with season, using variable discharge. Season 1 = autumn (SO); Season 2 = winter (NDJFM); Season 3 = spring (AM); Season 4 = summer (JJA).
18
Figure 13a and 13b show that in the forest-dominated catchments (C1 and C2) and the silt underlain catchment (C14) that the export of DOC is overestimated by the uniform specific discharge during all seasons. In winter these differences were up to 50%. In the wetland-dominated catchments the export was underestimated during all seasons, but mainly during spring by up to 43%.
Figure 13. a) Difference in DOC export for autumn and winter, if variable discharge is used compared to uniform discharge. b) Difference in DOC export for spring and summer, if variable discharge is used compared to uniform discharge. The catchments are sorted according to how much of the catchment is covered by wetland, with highest areal coverage to the left and lowest to the right. Positive numbers in the diagram mean that using the assumption of uniform discharge the export is underestimated in this catchment and negative numbers mean the opposite.
The assumption of uniform discharge underestimates the export of DOC, S and Mg mainly in April, by 39-41% .In May the export rates for both the uniform and variable specific discharge are the same for these elements. For Ca the pattern is reversed, in April the assumption uniform discharge overestimates the export by 9% and in May by 38% (Figure 14).
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50
4 5 10 6 7 12 15 9 13 20 16 14 1 2
Difference in DOC export (%)
Catchment
Autumn (SO) Winter (NDJFM)
-50 -40 -30 -20 -10 0 10 20 30 40 50
4 5 10 6 7 12 15 9 13 20 16 14 1 2
Difference in DOC export (%)
Catchment
Spring (AM) Summer (JJA)
a)
b)
19
Figure 14. Difference in element export for the spring months, April and May, for a wetland catchment (C4), if variable discharge is used compared to uniform discharge. The catchments are sorted according to how much of the catchment is covered by wetland, with highest areal coverage to the left and lowest to the right. Positive numbers in the diagram mean that using the assumption of uniform discharge the export is underestimated in this catchment and negative numbers mean the opposite.
In the forest-dominated catchment (Figure 15) the export rates were overestimated using the uniform discharge for both spring months, for all elements. In April between 24 and 56% and in May between 11 and 23%.
Figure 15. Difference in element export for the spring months, April and May, for a forest catchment (C2), if variable discharge is used compared to uniform discharge. Positive numbers in the diagram mean that using the assumption of uniform discharge the export is underestimated in this catchment and negative numbers mean the opposite.
At the catchment outlet (Figure 16) the differences between the variable and the uniform discharge are less distinct, than in the other catchments. In April the export rates are underestimated by 6-12%
20
(DOC,S and Mg) and overestimated by 20% (Ca). In May the pattern is reversed with overestimation of DOC, S and Mg by 12-14% and underestimation of Ca export by 12%.
Figure 16. Difference in element export for the spring months, April and May, for the catchment outlet (C16), if variable discharge is used compared to uniform discharge. Positive numbers in the diagram mean that using the assumption of uniform discharge the export is underestimated in this catchment and negative numbers mean the opposite.
5.3 Result of discharge-concentration relationship
The b-parameter describing the discharge-DOC-concentration relationship is lowest in the wetland- dominated catchment (C4), which could speaks for overland flow and dilution. This factor increases in the forest-dominated catchment (C2) and is highest in the silt underlain catchment (C14) (Figure 17 a- c).
For Mg the pattern of the discharge-concentration relationship is different, the value is lowest in the wetland-dominated catchment (C4) and then increasing in the forest-dominated (C2) and the silt underlain catchment (C14). This means higher conductivity and/or concentration in the wetland- dominated catchment and lower conductivity and/or concentration in the silt underlain catchment (Figure 18 a-c), which speaks for a deeper source for this element. For DOC the values are almost all positive or close to positive, which indicates that sources are from shallower flow paths and/or leaching of DOC is limited to shallow soil layers. In spring and during the wettest year of the time series the b-parameter is slightly negative in wetland-dominated catchments, which could indicate overland flow. In the forest-dominated catchment the b value increases slightly from the driest to the wettest year, indicating less weathering (Table 2).
21
Figure 17. Log-log diagram of: a) Five years of DOC concentration in the stream water of a wetland-dominated catchment (C4). b) Five years of DOC concentration in the stream water of a forest-dominated catchment (C2).
c) Five years of DOC concentration in the stream water of a silt underlain catchment (C14).
y = 19,868x0,075 1
10 100
0,00 0,01 0,10 1,00 10,00 100,00
DOC (mg/L)
Q (mm/d)
y = 13,7x0,1412 1
10 100
0,00 0,01 0,10 1,00 10,00
DOC (mg/L)
Q (mm/d)
y = 29,792x-0,162 1
10 100
0,01 0,10 1,00 10,00
DOC (mg/L)
Q (mm/d)
22
Figure 18. Log-log diagram of: a) Five years of Mg concentration in the stream water of a wetland-dominated catchment (C4). b) Five years of Mg concentration in the stream water of a forest-dominated catchment (C2). c) Five years of Mg concentration in the stream water of a silt underlain catchment (C14).
The b-parameter is negative for the weathering derived Mg for the whole time series and landscapes, y = 0,3189x-0,374
0,01 0,1 1
0,01 0,10 1,00 10,00 100,00
Mg (mg/L)
Q (mm/d)
y = 0,6411x-0,085
0,1 1 10
0,00 0,01 0,10 1,00 10,00 100,00
Mg (mg/L)
Q (mm/d)
y = 0,8795x-0,093
0,1 1 10
0,00 0,01 0,10 1,00 10,00
Mg (mg/L)
Q (mm/d)
a)
b)
c)
23
Table 2. b-parameter for different time series and seasons.
DOC Mg
Whole time series (forest-dominated catchment) 0.08 -0.09 Whole time series (wetland-dominated catchment) -0.16 -0.37 Whole time series (silt underlain catchment) 0.14 -0.09 Driest year of the time series (forest-dominated catchment) 0.12 -0.11 Wettest year of the time series (forest-dominated catchment) 0.02 -0.09 Driest year of the time series (wetland-dominated catchment) -0.11 -0.31 Wettest year of the time series (wetland-dominated
catchment) -0.24 -0.42
Driest year of the time series (silt underlain catchment) 0.14 -0.09 Wettest year of the time series (silt underlain catchment) 0.12 -0.11
Spring (forest-dominated catchment) 0.14 -0.07
Summer (forest-dominated catchment) 0.07 -0.11
Spring (wetland-dominated catchment) -0.23 -0.42
Summer (wetland-dominated catchment) 0.04 -0.23
24
6. Discussion
The aim of this study was to compare if calculations of element export of C (in the form of DOC), S, Ca and Mg would differ using discharge from only one subcatchment (C7) on one hand and variable discharge, measured at each catchment outlet on the other hand. Using the assumption of uniform specific discharge instead of a spatially variable specific discharge results in significant overestimation of element export from the forest-dominated catchments (C1 and C2) and underestimation of export from wetland-dominated catchments (C4 and C5) (Figure 4 and 5). All elements show relatively similar over- and underestimation for the same catchments, which is surprising as the origin of these elements is fundamentally different. The main factor responsible for the differences between uniform and variable discharge seems to be the runoff generation in the subcatchments (Figure 6). The differences in flow per unit area are between +26% and -27%. This is slightly higher than in the study of Karlsen et al. (2016), with a variability of 19% on an annual scale, but the reference catchment in this study was the catchment outlet, C 16, and differences should be expected. Other landscape properties like stream order or catchment size do not appear to play a major role. Even between similar catchments, the effects can be very different. The wetland-dominated catchment C5 for example is located at a lake outlet that might affect the flow from this catchment and some of the DOC could for example have time to react with iron forming organic complexes, sediment in the lake and therefore lower the concentrations down streams.
The mean DOC export rates of 3530-11130 kg/km²a (Figure 7) for the whole time series are lower than in a previous study of the Krycklan catchment, 14 800-99 100 kg/km²a, (Ågren et al. 2007) and more in line with results of a catchment in Southern Finland, 6100 kg/km²a (Huotari et al. 2013).
Results of previous studies and the result this study match in other areas, for example that wetlands have the highest export rates (Ågren et al. 2007) and catchments underlain by silt have the lowest export rates (Ågren et al. 2014) (Figure 11). Small headwaters (with stream order 1), including C1, C2 and C4-6, have in most cases (4 out of 5) the highest DOC export per unit area. The export of the other elements S, Ca and Mg are 137-794, 331-1733 and 119-403 kg/km²a, respectively, for the different subcatchments in the Krycklan catchment during the whole five years time series. Those values are within or slightly higher than seen in previous studies with export rates of 180-750 and 67-260 kg/km²a for Ca and Mg respectively in the Krycklan catchment (Lidman et al. 2014), but lower than in the Siberian catchment with 3000 and 540 kg/km²a respectively (Pokrovsky et al. 2006). The export of sulphur is also similar or higher compared to previous studies in the area with 180-390 kg/km²a (Mörth et al. 2008). An exception is C20 which has about twice as much element export of S, Ca and Mg, than reported in previous studies. The reason for the export of these elements being as large in this catchment is unclear, because the catchment properties, size, vegetation, soil, stream order, are similar to other subcatchments, but with much lower export rates (Figure 8). On the hand is information on this particular subcatchment sparse and the gauging station was most likely established
25
less than ten years ago and discharge and stream concentration data from this catchment are probably not included in previous studies.
The over- and underestimation of the element export is fairly constant through the seasons, in the sense that if the export from one catchment is overestimated in winter it was, in most cases, also overestimated in other seasons (Figures 13a and b). Although the per percent over- and underestimation in the catchments might be similar through the seasons the differences in the total export of element differs greatly during the seasons. In spring the export rates increase by up to one order-of-magnitude and during this season small differences between uniform and variable specific discharge could have a large impact (Figures 11 and 12). The comparison between the driest and the wettest year in the time series does not show any apparent effect on the export of sulphur, calcium and magnesium, whereas export of DOC was lowest during the driest year and highest during the wettest year (Figures 9a and b and 10a and b). The evaluation of the b-parameter shows that overland flow and dilution is more common in wetland-dominated catchments. Changes in depths of the flowpaths seem to affect the DOC leaching more than the weathering of for example Mg (see figures 17a-c, 18a- c and table 2).
A very important factor is obviously which catchment is used as the reference for the uniform discharge. In this study C7 was used, because this measuring station was established in 1980 and continuous time series exist since then, this means that many studies used data from this particular catchment and which makes it possible to compare the data more easily. Another option would have been the catchment outlet, C16, this way the results would most likely have been different, but significant differences between uniform and variable specific discharge would still have emerged.
If in a future climate the precipitation will likely increase, runoff generation might increase as well and as a consequence the element export, however, these processes are very complex. How to deal with the implications of this study is difficult to answer. Variable specific discharge should of course be preferred, but increasing the number of discharge measurement and water analysis significantly seems impossible, because that would be too expensive and time consuming. One possible solution could be the subdivision of larger catchments into smaller units based on landscape feature. More studies, both in boreal and other landscapes, are needed to confirm the results of this study.
26
7. Conclusions
The aim of this study was to investigate if the element export is similar using variable discharge or the assumption of uniform specific discharge. The following conclusions have been made:
- Uniform discharge underestimates the export of elements from wetland-dominated catchments by up to 26% and overestimates the export from forest-dominated catchments by up to 30%.
- Discharge and element export per unit area from subcatchments is markedly different than for the whole catchment.
- Differences are strongly connected to discharge generation from the subcatchments.
27
8. Acknowledgments
First of all I would like to thank my supervisor Thomas Grabs for his invaluable help. Patiently reading through my drafts, answering a gazillion questions and keeping me on the right track. I would also like to thank Kevin Bishop for giving me opportunity to work on this thesis. Further I would like to thank Reinert Karlsen and Ida Taberman (SLU) for giving me accesses to all the data I needed and all the hard-working people of the Krycklan catchment study for collecting the data, even in the middle of winter.
28
9. References
Ågren, A., Buffam, I., Jansson, M. & Laudon, H. (2007) Importance of seasonality and small streams for the landscape regulation of dissolved organic carbon export. Journal of Geophysical Research, vol.
112, doi:10.1029/2006JG000381.
Ågren, A.M., Buffam, I., Cooper, D.M., Tiwari, T., Evans ,C.D. & Laudon, H. (2014) Can the heterogeneity in stream dissolved organic carbon be explained by contributing landscape elements?
Biogeosciences, vol. 11(4), pp. 1199–1213.
Björkvald, L., Buffam, I., Laudon H. & Mörth, C.-M. (2008) Hydrogeochemistry of Fe and Mn in small boreal streams: The role of seasonality, landscape type and scale. Geochimica et Cosmochimica Acta, vol. 72(12), pp. 2789–2804.
Buffam, I., Laudon, H., Temnerud, J., Mörth, C.M. & Bishop, K. (2007) Landscape-scale variability of acidity and dissolved organic carbon during spring flood in a boreal stream network. Journal of Geophysical Research, vol. 112, doi:10.1029/2006JG000218.
Forsum, Å., Laudon, H. & Nordin, A. (2008) Nitrogen uptake by Hylocomium splendens during snowmelt in a boreal forest. Ecoscience, vol. 15(3), pp. 315–319.
Huotari, J., Nykänen, H., Forsius, M. & Arvola, L. (2013) Effect of catchment characteristics on aquatic carbon export from a boreal catchment and its importance in regional carbon cycling. Global Change Biology, vol. 19(12), pp. 3607–3620.
Karlsen, R. H., Seibert, J. Grabs, T., Laudon, H., Blomkvist, P. & Bishop, K. (2016). The assumption of uniform specific discharge: unsafe at Any time? Hydrological Processes, (In press)
doi:10.1002/hyp.10877
Kjellström, E., Abrahamsson, R., Boberg, P., Jernbäcker, E., Karlberg, M., Morel, J. & Sjöström, Å.
(2014). Klimatförändring i Sverige och världen. Swedish Meteorological and Hydrological and Institute (SMHI). Accessible at: http://www.smhi.se/forskning/forskningsnyheter/klimatforandring-i- sverige-och-varlden-i-ny-svensk-rapport-1.81497 [Accessed 6-6-2016]
Kottek, M., Grieser, J., Beck, C., Rudolf, B. & Rubel, F. (2006) World Map of the Köppen-Geiger climate classification updated. Meteorologische Zeitschrift, vol. 15(3), pp. 259–263.
Laudon, H., Sjöblom, V., Buffam, I., Seiber,t J. & Mörth M. (2007) The role of catchment scale and landscape characteristics for runoff generation of boreal streams. Journal of Hydrology, vol. 344(3–4), pp. 198–209.
Laudon, H., Berggren, M., Ågren, A., Buffam, I., Bishop, K., Grabs, T., Jansson, M. & Köhler, S.
(2011) Patterns and dynamics of dissolved organic carbon (DOC) in boreal streams: The role of processes, connectivity, and scaling. Ecosystems, vol. 14(6), pp. 880–893.
Laudon, H., Taberman, I., Ågren, A., Futter, M., Ottosson-Löfvenius, M. & Bishop, K. (2013) The Krycklan Catchment Study-A flagship infrastructure for hydrology, biogeochemistry, and climate research in the boreal landscape. Water Resources Research, vol. 49(10), pp. 7154–7158.
Ledesma, J.L.J., Grabs, T., Bishop, K., Schiff, S. & Köhler S.J. (2015) Potential for long-term transfer of dissolved organic carbon from riparian zones to streams in boreal catchments. Global Change Biology, vol. 21(8), pp. 2963–2979.
29
Lidman, F., Köhler, S.J., Mörth, C.-M. & Laudon, H. (2014) Metal transport in the boreal landscape:
The role of wetlands and the affinity for organic matter. Environmental Science & Technology, vol.
48(7), pp. 3783–3790.
Lyon, S.W., Nathanson, M., Spans, A., Grabs, T., Laudon, H., Temnerud, J., Bishop, K.H. & Seibert, J. (2012) Specific discharge variability in a boreal landscape. Water Resources Research, vol. 48(8), doi:10.1029/2011WR011073.
Mörth, C.-M., Laudon, H., Mellqvist, E., Torssander, P. & Giesler, R. (2008) Sources of stream water sulfate during the spring snowmelt in boreal streams: Evidence from δ34S isotope measurements.
Journal of Geophysical Research, 113(G01005), doi:10.1029/2007JG000457.
Nadeau, T.-L. & Rains, M.C. (2007) Hydrological connectivity between headwater streams and downstream waters: How science can inform policy. Journal of the American Water Resources Association, vol. 43(1), pp. 118–133.
Pokrovsky, O. S., Schott, J. & Dupré, B. (2006). Trace element fractionation and transport in boreal rivers and soil porewaters of permafrost-dominated basaltic terrain in Central Siberia. Geochimica et Cosmochimica Acta, vol. 70(13), pp. 3239–3260.
Rember, R.D. & Trefry, J.H. (2004) Increased concentrations of dissolved trace metals and organic carbon during snowmelt in rivers of the Alaskan arctic. Geochimica et Cosmochimica Acta, vol. 68(3), pp. 477–489.
Seibert, J., Grabs, T., Köhler, S., Laudon, H., Winterdahl, M. & Bishop, K. (2009) Linking soil-and stream-water chemistry based on a riparian flow-concentration integration model. Hydrology and Earth System Sciences, vol. 13(12), pp. 2287–2297.
Laudon, H. (2016) The Krycklan Catchment Study: A unique infrastructure for field based research on hydrology, biogeochemistry and ecology, a Hitcher’s Guide, v. 3.0. Swedish University of
Agricultural Sciences (SLU). Available at:
https://www.slu.se/globalassets/ew/org/inst/esf/forsoksparker/vindeln/krycklan/krycklan-field- guide2.pdf [Accessed 6-6-2016]
SMHI (2015) Sveriges klimat (online)
Available at: http://www.smhi.se/kunskapsbanken/klimat/sveriges-klimat-1.6867.[Accessed 6-6-2016]
Teutschbein, C., Grabs, T., Karlsen, R.H., Laudon, H. & Bishop K. (2015) Hydrological response to changing climate conditions: Spatial streamflow variability in the boreal region. Water Resources Research, vol. 51, pp. 9425–9446.
Tiwari, T., Laudon, H., Beven, K. & Ågren, A.M. (2014) Downstream changes in DOC: Inferring contributions in the face of model uncertainties. Water Resources Research, vol. 50(1), pp. 514–525.
Tranvik, L.J. & von Wachenfeldt, E. (2009) Interactions of dissolved organic matter and humic substances. In: Likens, G. E. (Ed.), Encyclopedia of Inland Waters, Oxford: Academic Press, pp. 754–
760.
Wastenson, L., Raab, B. & Vedin, H. (1995) Sveriges nationalatlas: Klimat, sjöar och vattendrag, Stockholm: Sveriges nationalatlas (SNA).
