Chemical oxygen demand in lakes
Summary
The chemical oxygen demand of inland freshwaters indicates the amount of organic matter in water bodies and is thus closely related to the humus content of the waters. The background concentration of humus is strongly influenced by natural factors such as the amount of peatlands in the drainage areas, therefore the development of this indicator has separately considered lakes with low humus and humic lakes. Lakes with low humus are naturally clearer, and humic lakes are darker. The chemical oxygen demand in inland freshwaters is currently at a higher level than at the turn of the 21st century, when it was on average at its lowest during the monitoring period. However, the chemical oxygen demand is currently at a lower level than in the early part of the monitoring period in the 1960s, when inland freshwaters were still subjected to intense point-source pollution, particularly from municipal wastewater and the chemical forestry industry. The natural level of chemical oxygen demand, independent of human activity, is unknown, which complicates the assessment of human impacts.
Status
Compared to the good conditions around the turn of the 21st century, the current state of the indicator is satisfactory (see the “Information about the service” page). With ninety percent certainty, the chemical oxygen demand (COD) in lakes is at least about 16 percent higher than during the period 1999-2003, when it was at its lowest in most lakes during the monitoring period.
Historically, however, the COD may have been lower than during the turn of the 21st century, as systematically collected data is only available from most lakes starting from the 1960s. At that time, the load on water bodies was already elevated above natural levels, particularly due to inadequate treatment of wastewater and the chemical forestry industry. The estimated condition refers to the average state in a sample of various different lakes. The current states of individual lakes in relation to the benchmark period around the turn of the 21st century may vary significantly, and some lakes may also have had periods when the COD was lower than at the turn of the 21st century.
Trend
The development of the indicator is deteriorating (see the “Information about the service” page). The COD has increased during the 21st century in both clear, low-humus and darker, humic lakes, with ninety percent certainty by at least 0.7 percent per year.
The recent increase in COD is likely explained particularly by the growing humus load caused by the drainage of peatlands, which climate change probably exacerbates (Asmala et al. 2019, Lepistö et al. 2021, Nieminen et al. 2021, Vilmi et al. 2021). The effects of drainages and climate change likely amplify each other. Due to drainages and climate change, both primary production and the decomposition and leaching of organic matter during the period of unforzen soil have increased, which in turn raises the COD in lakes.
This recent increasing trend was preceded by a corresponding rate of decrease in the COD, which continued fairly steadily from the early 1960s to the beginning of the 21st century. The decreasing trend in the earlier part of the monitoring period was particularly influenced by the advancement in wastewater treatment, but also by the acidification of water bodies and soil due to industrial sulfur emissions, which continued until the 1980s and 1990s.
Significance
The indicator describes the condition of lakes and, more precisely, the load caused by organic matter within them. Therefore, an increase in the value of the indicator is negatively associated with the condition of the water body. COD reacts to all organic loading and may decrease as a result of acidification, but it is especially related to water color and humus content.
COD has been monitored in the largest Finnish lakes using methods that are comparable with each other from the early 1960s to the present day (see “Data used”). However, a challenge with the data is that human-induced loading on inland waters had already significantly affected the condition of the inland waters before the monitoring began, so the data does not directly provide a current state of the surface waters compared to the historical completely natural state.
COD is naturally higher in areas dominated by peatlands and in small, shallow headwater lakes because these naturally have high humus content and short water residence time. In large lakes, on the other hand, water generally stays longer, which allows for the precipitation and accumulation of humus at the lake bottom.
In addition to natural background leaching, the runoff of organic matter in Finland has particularly increased due to extensive drainage of peatlands. Dark water can absorb more heat than clear water, which is detrimental to species that live in cold water, such as salmonids.
References
Data used
The monitoring of the condition of surface waters forms one of Finland’s most extensive datasets describing the state of nature and the environment. The systematic monitoring of surface water quality began in the 1960s, and especially since 2009, monitoring has increased under the requirements of the European Union’s Water Framework Directive. Currently, the water quality registry has accumulated over 75,000 observation sites, of which more than two hundred can form long time series, with observations collected at least annually.
Long-term monitoring also allows for the classification of lakes into naturally low-humic, humic, and highly-humic, enabling the examination of developmental trajectories separately in different types of water bodies. Naturally highly-humic lakes are few, and they have also accumulated less monitoring data, which is why they have been excluded from the analysis in this indicator.
The indicator includes the following 28 low-humic lakes (in parentheses is the province where located):
Höytiäinen (North Karelia)
Inarijärvi (Lapland)
Juojärvi (North Savo, North Karelia)
Keitele (Central Finland, North Savo)
Kermajärvi (North Karelia)
Kivijärvi (Central Finland)
Kivijärvi (South Karelia)
Kolima (Central Finland)
Konnevesi (Central Finland, North Savo)
Kukkia (Pirkanmaa, Kanta-Häme)
Kuolimo (South Savo, South Karelia)
Leppävesi (Central Finland)
Mallasvesi (Pirkanmaa)
Muojärvi (North Ostrobothnia)
Puruvesi (South Savo, North Karelia)
Puula (South Savo, Central Finland)
Pyhäjärvi (Central Finland)
Pyhäjärvi (South and North Karelia)
Pyhäjärvi (Satakunta and Southwest Finland)
Pyhäjärvi (North Ostrobothnia)
Pyhäjärvi (Kymenlaakso)
Päijänne (Central Finland, Pirkanmaa, Päijät-Häme)
Roine (Pirkanmaa)
Saimaa (South Karelia, South Savo)
Suvasvesi (North Savo, North Karelia)
Viinijärvi (North Karelia)
Vuohijärvi (Kymenlaakso, South Savo)
Yli-Kitka (Lapland, North Ostrobothnia)
The indicator includes the following 25 humic lakes (in parentheses is the province where located):
Haukivesi (South and North Savo)
Kallavesi (North Savo)
Kemijärvi (Lapland)
Keurusselkä (Central Finland, Pirkanmaa)
Kiantajärvi (Kainuu)
Koitere (North Karelia)
Kyyvesi (South Savo)
Lappajärvi (South Ostrobothnia)
Lentua (Kainuu)
Lestijärvi (Central Ostrobothnia)
Näsijärvi (Pirkanmaa)
Ontojärvi (Kainuu)
Oulujärvi (North Ostrobothnia, Kainuu)
Orivesi (North Karelia, South Savo)
Pielinen (North Karelia)
Pihlajavesi (South Savo)
Punelia (Kanta-Häme)
Pyhäjärvi (Pirkanmaa)
Pyhäselkä (North Karelia)
Simojärvi (Lapland)
Sääksjärvi (Satakunta)
Unnukka (North Savo)
Vanajavesi (Pirkanmaa, Kanta-Häme)
Vuotjärvi (North Savo)
Ähtärinjärvi (South Ostrobothnia)
More information about the data:
Indicator calculation
The indicator includes samples taken from the beginning of 1960, collected at a depth of 1-2 meters and measured for chemical oxygen demand (CODMn) using permanganate ion as the oxidizing agent.
Since the goal of the indicator is to express a generalizable state and developmental trajectory across several very different lakes, the calculations aim to account for differences between lake types, individual lakes, different basins within lakes, and even individual sampling sites. In practice, this is achieved using so-called hierarchical generalized statistical mixed models, which can accurately consider the structure of the data, ensuring that individual extreme observations or sampling sites do not significantly distort the generalizations produced by the model. These statistical models have been utilized both to determine the state and developmental direction of the indicator and to produce the indicator graph.
To define the state of the indicator, the data first identified a five-year comparison window during which the chemical oxygen demand of most lakes was at its lowest. The chemical oxygen demand of the last five years has then been compared to this baseline period, and the average relative difference between the current state and the baseline period has been estimated. More detailed information on how the state of the indicator is classified can be found in the “Information about the service” page.
The development direction has been assessed from the average trend from the beginning of the 21st century to the present moment. More detailed information on how the development of the indicator is classified can be found in the “Information about the service” page. For a longer-term perspective, a trend has also been calculated comparing the start of the time series to the year 2000.
Additional information on the calculation of the indicator:
Ask for further information
Sari Mitikka
Senior research scientist (Syke), Surface water quality monitoring
Mika Nieminen
Principal scientist (Luke), Nutrient cycles and water loading