Time filter

Source Type

Grigorievskiy A.,Aalto University | Miche Y.,Aalto University | Ventela A.-M.,Pyhajarvi Institute | Severin E.,Lille University of Science and Technology | And 3 more authors.
Neural Networks | Year: 2014

In this paper, an Optimally Pruned Extreme Learning Machine (OP-ELM) is applied to the problem of long-term time series prediction. Three known strategies for the long-term time series prediction i.e. Recursive, Direct and DirRec are considered in combination with OP-ELM and compared with a baseline linear least squares model and Least-Squares Support Vector Machines (LS-SVM). Among these three strategies DirRec is the most time consuming and its usage with nonlinear models like LS-SVM, where several hyperparameters need to be adjusted, leads to relatively heavy computations. It is shown that OP-ELM, being also a nonlinear model, allows reasonable computational time for the DirRec strategy. In all our experiments, except one, OP-ELM with DirRec strategy outperforms the linear model with any strategy. In contrast to the proposed algorithm, LS-SVM behaves unstably without variable selection. It is also shown that there is no superior strategy for OP-ELM: any of three can be the best. In addition, the prediction accuracy of an ensemble of OP-ELM is studied and it is shown that averaging predictions of the ensemble can improve the accuracy (Mean Square Error) dramatically. © 2013 Elsevier Ltd.

Nurnberg G.K.,Freshwater Research | Tarvainen M.,Pyhajarvi Institute | Tarvainen M.,University of Turku | Ventela A.-M.,Pyhajarvi Institute | Sarvala J.,University of Turku
Inland Waters | Year: 2012

We quantified internal phosphorus (P) load for 26 years in the polymictic, large (155 km2) and shallow (mean depth 5.5 m) Lake Säkylän Pyhäjärvi, which was heavily biomanipulated by fish removal. Internal load was estimated as (1) partially net estimates from in situ P summer increases, (2) net estimates from P budgets (mass balance approach), and (3) gross estimates from predicted active sediment release area and sediment P release, dependent on August lake temperature. Long-term averages of these estimates were similar and large at about 60% of average external load (105 mg m-2 yr-1) and were larger than external load in years with high water temperature and low water load. Regression analysis revealed that external load is decreasing but internal load is increasing over time. Internal load was negatively correlated with annual water load and positively correlated with lake water temperature. Long-term average annual or summer P concentrations are adequately predicted by a P mass balance model that includes external load, internal load (Method 3), and sedimentation as independently predicted retention. Predictability was poor for individual years, however, partially due to the poor correlation of observed lake and outflow P concentrations, the variable abundance of planktivorous fish, and reflecting the violation of the steady state assumption when individual years are modeled. Scenario modeling shows that biomanipulation cancels out the effects of internal load and forecasts a rapid increase of internal load and P concentration due to climate change; therefore, measures that further decrease external and internal P load and strengthen biomanipulation are recommended. © International Society of Limnology 2012.

Patynen A.,Finnish Environment Institute | Patynen A.,University of Jyväskylä | Elliott J.A.,UK Center for Ecology and Hydrology | Kiuru P.,Finnish Environment Institute | And 3 more authors.
Boreal Environment Research | Year: 2014

We linked the models PROTECH and MyLake to test potential impacts of climate-change-induced warming on the phytoplankton community of Pyhäjärvi, a lake in southwest Finland. First, we calibrated the models for the present conditions, which revealed an apparent high significance of internal nutrient loading for Pyhäjärvi. We then estimated the effect of two climate change scenarios on lake water temperatures and ice cover duration with MyLake. Finally, we used those outputs to drive PROTECH to predict the resultant phy-toplankton community. It was evident that cyanobacteria will grow significantly better in warmer water, especially in the summer. Even if phosphorus and nitrogen loads to the lake remain the same and there is little change in the total chlorophyll a concentrations, a higher proportion of the phytoplankton community could be dominated by cyanobacteria. The model outputs provided no clear evidence that earlier ice break would advance the timing of the diatom spring bloom. © 2014.

Gonzales-Inca C.A.,University of Turku | Kalliola R.,University of Turku | Kirkkala T.,Pyhajarvi Institute | Lepisto A.,Finnish Environment Institute
Water Resources Management | Year: 2015

Finnish arable land is typically located on flat areas, where the fields are mostly drained and sub-drained to control the water tables. These areas are highly susceptible to nutrient loss, which affects the water quality of rivers and lakes. Therefore, it is very important to understand current landscape pattern and processes controlling water quality, not only identifying factors affecting it, but also identifying strategies and restoring areas for mitigation. We studied linkage of 21 years (1990–2011) of water quality (WQ) data from 16 agricultural watersheds, using landscape indices at three functional scales: watershed-wide, saturation-excess zone, and riparian zone (of varying widths). The hydro-biogeochemical functional areas of watershed were obtained by digital terrain analysis. Statistical analyses by generalized linear model and multivariate redundancy analysis indicated that the fraction of watershed in agricultural use was linked to most of the studied water quality variables. The relationships varied across the seasons: they were strongest during high flow periods (spring and autumn) when also highest nutrient losses occur. Total suspended sediment concentrations were linked to critical source areas. Riparian vegetation index was important explaining nitrate concentrations in autumn. Terrain-based mapping of hydro-biogeochemical functional areas provides a rapid identification of potential sites to mitigate diffuse nutrient pollution, particularly in riparian zones. © 2014, Springer Science+Business Media Dordrecht.

Ventela A.-M.,Pyhajarvi Institute | Kirkkala T.,Pyhajarvi Institute | Lendasse A.,Aalto University | Tarvainen M.,Pyhajarvi Institute | And 2 more authors.
Hydrobiologia | Year: 2011

Säkylän Pyhäjärvi (SW Finland) is an example of a large and shallow lake suffering from eutrophication. During the last 20 years, the quality and general usability of water in Pyhäjärvi have shown considerable variation driven by both a variety of human activities and climate-related factors such as wet and dry years. The lake has been thoroughly studied for decades and has been the object of comprehensive restoration activities both in the catchment and in the lake since the 1990s. Large variety of water protection measures like wetlands, sedimentation ponds and filtering systems have been implemented in the catchment area to reduce external nutrient load. Another important tool for Pyhäjärvi restoration is biomanipulation, done by local commercial fishermen in winter. Twenty-five percent of the annual phosphorus input is removed with fish catch. Currently, restoration work is facing new challenges: short or even nearly missing ice cover period and increased winter time nutrient load from the catchment. In the 2000s, there were 3 years with exceptionally short ice period, allowing only brief winter seining periods. Consequently, the biomanipulation catch was very low in 2007 and 2008 leading to observable food web effects. Phosphorus load was especially high in winters 2006/2007 and 2008/2009. On the basis of the data of 1987-2008, we have tested the hypothesis if climate-related winter time variables like phosphorus load, air temperature and precipitation would affect the water quality of the lake in following summer, here measured as chlorophyll a concentration in the lake water. A linear model has been used and a validation procedure has been performed to select the best variables. Our results indicate some of the linkages between climate-related catchment processes and the ecological status of the lake. © Springer Science+Business Media B.V. 2010.

PubMed | Pyhajarvi Institute, CAS Nanjing Institute of Geography and Limnology and University of Helsinki
Type: Journal Article | Journal: PloS one | Year: 2015

A yearlong campaign to examine sediment resuspension was conducted in large, shallow and eutrophic Lake Taihu, China, to investigate the influence of vegetation on sediment resuspension and its nutrient effects. The study was conducted at 6 sites located in both phytoplankton-dominated zone and macrophyte-dominated zone of the lake, lasting for a total of 13 months, with collections made at two-week intervals. Sediment resuspension in Taihu, with a two-week high average rate of 1771 gm(-2)d(-1) and a yearly average rate of 377 gm(-2)d(-1), is much stronger than in many other lakes worldwide, as Taihu is quite shallow and contains a long fetch. The occurrence of macrophytes, however, provided quite strong abatement of sediment resuspension, which may reduce the sediment resuspension rate up to 29-fold. The contribution of nitrogen and phosphorus to the water column from sediment resuspension was estimated as 0.34 mgL(-1) and 0.051 mgL(-1) in the phytoplankton-dominated zone. Sediment resuspension also largely reduced transparency and then stimulated phytoplankton growth. Therefore, sediment resuspension may be one of the most important factors delaying the recovery of eutrophic Lake Taihu, and the influence of sediment resuspension on water quality must also be taken into account by the lake managers when they determine the restoration target.

Koskiaho J.,Finnish Environment Institute | Lepisto A.,Finnish Environment Institute | Tattari S.,Finnish Environment Institute | Kirkkala T.,Pyhajarvi Institute
Water Science and Technology | Year: 2010

Automatic on-line measurement stations for water quality components and water level were equipped with dataloggers and GSM transmitters; the stations were installed at two sites in the Yläneenjoki river basin, SW Finland. Measurements during five seasons in 2006-2007 were conducted to find out whether the produced data would provide more accurate estimates of material and nutrient transport than traditional water sampling. Sensor-based monitoring estimates for transport of total suspended solids (TSS) were clearly higher (difference -6-61%), total phosphorus also higher, and that of nitrate (NO 3-N) somewhat lower (difference (-51%-4%), as compared with manual sampling based estimates. The winter season studied here was mild i.e. winter-type which is becoming more common in Finland with the changing climate. Sensor-based monitoring proved its benefits particularly in such conditions. © IWA Publishing 2010.

Kirkkala T.,Pyhajarvi Institute | Ventela A.-M.,Pyhajarvi Institute | Tarvainen M.,Pyhajarvi Institute
Agricultural and Food Science | Year: 2012

Diffuse load mitigation is a prevailing global challenge due to the eutrophication of water bodies. We report here long-term nutrient removal performance of two on-site sand filters (F1 and F2) in southwestern Finland, established in the 1990s. The sand filters were enhanced with a layer of phosphorus binding material Fosfilt-s, a side product of titanium dioxide production. The monitoring periods were 4.5 and 3.5 years for F1 and F2, respectively. F1 only worked for some months due to blockage of the crushed stone layer. After renovation (1999), the filter worked for a year but then the Fosfilt-s layer clogged and the filter increased suspended solids (SS) and dissolved reactive phosphorus (DRP) loads by 36% and 19% on average, respectively. Total P (PTOT) load was not affected. The structure and performance of F2 were more successful due to a better water distribution system. On average, 61% of the inflowing suspended solids, 37% of the PTOT and 45% of the DRP were removed during the monitoring period. We conclude that these filter types have the potential to be further developed into useful tools for nutrient load reduction. Development work should be focused on the treatment of subsurface drainage waters with high concentrations of dissolved nutrients, in particular. Long term field data is needed, because laboratory tests cannot fully simulate natural circumstances.

Verta M.,Finnish Environment Institute | Salo S.,Finnish Environment Institute | Korhonen M.,Finnish Environment Institute | Porvari P.,Finnish Environment Institute | And 2 more authors.
Science of the Total Environment | Year: 2010

We conducted a whole-lake experiment by manipulating the stratification pattern (thermocline depth) of a small polyhumic, boreal lake (Halsjärvi) in southern Finland and studying the impacts on lake mercury chemistry. The experimental lake was compared to a nearby reference site (Valkea-Kotinen Lake). During the first phase of the experiment the thermocline of Halsjärvi was lowered in order to simulate the estimated increase in wind speed and in total lake heat content (high-change climate scenario). The rate of methyl mercury (MeHg) production during summer stagnation (May-August) was calculated from water profiles before the treatment (2004), during treatment (2005, 2006) and after treatment (2007). We also calculated fluxes of MeHg from the epilimnion and from the hypolimnion to the sediments using sediment traps. Experimental mixing with a submerged propeller caused a 1.5-2. m deepening of the thermocline and oxycline. Methyl mercury production occurred mostly in the oxygen free layers in both lakes. In the experimental lake there was no net increase in MeHg during the experiment and following year; whereas the reference lake showed net production for all years. We conclude that the new exposed epilimnetic sediments caused by a lowering of the thermocline were a major sink for MeHg in the epilimnion. The results demonstrate that in-lake MeHg production can be manipulated in small lakes with anoxic hypolimnia during summer. The climate change induced changes in small boreal lakes most probably affect methyl mercury production and depend on the lake characteristics and stratification pattern. The results support the hypothesis that possible oxygen related changes caused by climate change are more important than possible temperature changes in small polyhumic lakes with regularly occurring oxygen deficiency in the hypolimnion. © 2010 Elsevier B.V.

Kirkkala T.,Pyhajarvi Institute | Ventela A.-M.,Pyhajarvi Institute | Tarvainen M.,Pyhajarvi Institute
Journal of Environmental Quality | Year: 2012

The River Yläneenjoki catchment in southwest Finland is an area with a high agricultural nutrient load. We report here on the nutrient removal performance of three on-site lime-sand filters (F1, F2, and F3), established within or on the edge of the buffer zones. The filters contain burnt lime (CaO) or spent lime [CaO, Ca(OH) 2, and CaCO 3]. Easily soluble lime results in a high pH level (>11) and leads to an efficient precipitation of soluble phosphorus (P) from the runoff. Water samples were taken from the inflow and outflow of each site in different hydrological situations. The length of the monitoring period was 4 yr for F1, 6 yr for F2, and 1.5 yr for F3. F1 and F2 significantly reduced the suspended solids (SS), total P (PTOT), and dissolved reactive P (DRP) in the treated water. The proportional reduction (%) varied but was usually clearly positive. Filter F3 was divided into two equal parts, one containing burnt lime and the other spent lime. Both filter parts removed PTOT and SS efficiently from the water; the burnt-lime part also removed DRP. The mixed-lime part removed DRP for a year, but then the efficiency decreased. The effect of filters on nitrogen compounds varied. We conclude that sand filters incorporating lime can be used together with buffer zones to reduce both P and SS load to watercourses. © 2012 by the American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America.

Loading Pyhajarvi Institute collaborators
Loading Pyhajarvi Institute collaborators