Control Systems and Computers

https://doi.org/10.15407/csc.2020.02.066

Control Systems and Computers, 2020, Issue 2 (286), pp. 66-76.

UDC 004.94

Halyna A. Pidnebesna, junior research scientist, Department for Technologies of Inductive Modelling, International Research and Training Center for Information Technologies and Systems of the NAS and MES of Ukraine, Glushkov Ave., 40, Kyiv, 03187, Ukraine, pidnnebesna@ukr.net

Bioproductivity of Dnieper Reservoirs 
Analysis by Inductive Methods

Introduction. Investigations of the laws of functioning of aquatic ecosystems, the influence of natural factors on the process of formation of biological productivity of reservoirs under anthropogenic loading are extremely important for the assessment of ecological safety and maintenance of ecological balance, and, as a consequence, for the development of scientifically sound water quality management methods and forecasts.

Goal. The task was to determine the factors that have the most significant influence on the state of water in the Dnieper reservoirs by constructing a model of dependence of the concentration of chlorophyll a in phytoplankton according to long-term observations in Kremenchug and Kakhovka reservoirs. The results of observations of the Institute of Sciences in 1976-1993.

Methods. The small amount of observational data and measurement errors make it difficult to solve the problem. Various inductive methods were used to obtain a satisfactory result. Algorithms are simulated: linear regression of LR, LASSO, combinatorial algorithm of MSU COMBI, and correlation algorithm for analysis of CAR factor rating. The coefficient of determination R2 and the corresponding multiple correlation coefficient R. were used to evaluate the adequacy of the models obtained.

Results. For the Kremenchug reservoir, models constructed using LASSO and COMBI were found to have a negative value for the coefficient of determination R2, that is, insufficient adequacy.

The model obtained by linear regression LR has a coefficient of determination of R2 = 0.204 (respectively, multiple correlation R = 0.452). This means that the model has satisfactory adequacy. But at the same time it has in its composition all the factors, that is, it does not select the most important ones.

The model obtained using the correlation algorithm for calculating the CAR regression rating has a coefficient of determination of R2 = 0.273 (respectively, multiple correlation R = 0.522), ie the model has a good degree of adequacy.

For the Kakhovka reservoir, the models obtained by the applied methods have a high degree of adequacy (RLR = 0.838, RCOMBI = 0.820, RCAR = 0.812, RLASSO = 0.803). The models also have similar structures.

Conclusion. The analysis of the obtained results showed that in all models obtained by different methods, for both reservoirs there is a factor – mineral form of nitrogen. This is natural, when eutrophying reservoirs nitrogen content becomes a factor that limits the development of phytoplankton. Three of the models obtained for the Kakhovka reservoir and two Kremenchug models with satisfactory adequacy include water runoff.

The conducted research suggests that these factors (mineral form of nitrogen and volume of water runoff) may be the most influential for the life of phytoplankton and determine the functioning of the Kakhovsky and Kremenchug reservoirs. The above refers to the named Dnieper reservoirs and an available sample of summer season statistics, which is characterized by intense “flowering” of water by blue-green algae.

 Download full text! (On English)

Keywords: inductive modeling methods, correlation algorithm CAR with calculation of regressors rating, combinatorial algorithm GMDH COMBI, LASSO, phytoplankton, chlorophyll a concentration.

1. Yang, X., Wu, X., Hao, H., Zhen-li, He., 2008. “Mechanisms and assessment of water eutrophication,” J. Zhejiang Univ. Sci B, 9v(3), pp. 197-209.
   https://doi.org/10.1631/jzus.B0710626
2. Raike, A., Pietilainen, O.P., Rekolainen, S., et al., 2003. “Trends of phosphorus, nitrogen, and chlorophyll a concentrations in Finnish rivers and lakes in 1975-2000,” The Science of the Total Environment, vol. 310, pp. 47-59.
   https://doi.org/10.1016/S0048-9697(02)00622-8
3. Worsfold, P.J., Monbet, P., Tappin, A.D., Fitzsimons, M.F., Stiles, D.A., McKelvie, I.D., 2008. “Characterisation and quantification of organic phosphorus and organic nitrogen components in aquatic systems: a review,” Anal. Chim. Acta. Epub, 624 (1), pp. 37-58.
   https://doi.org/10.1016/j.aca.2008.06.016
4. Elser, J.J., Bracken, M.E., Cleland, E.E., Gruner, D.S., Harpole, W.S., Hillebrand, H., Ngai, J.T., Seabloom, E.W., Shurin, J.B., Smith, J.E., 2007. “Global analysis of nitrogen and phosphorus limitation of primary producers in freshwater, marine and terrestrial ecosystems,” Ecol. Lett., 10 (12), pp. 1135-1142.
   https://doi.org/10.1111/j.1461-0248.2007.01113.x
5. Kureishevich, A.V., Nezbrytskaya, I.N., Stanislavchuk, A.V., 2016. “Activity of Antioxidant Enzymes of Cyanoprokaryota and Green Microalgae Cultures under Different Temperatures Conditions,” International Journal on Algae, 18 (2), pp. 169-177.
   https://doi.org/10.1615/InterJAlgae.v18.i2.60
6. Toxic Cyanobacteria in Water: A guide to their public health consequences, monitoring and management. [online] Available at: <https://www.who.int/water_sanitation_health/
   resourcesquality/toxcyanbegin.pdf> [Accessed 10 Jan. 2020].
7. Kureishevich, A.V., Pidnebesna, H.A., Stepashko, V.S., 2015. “Analysis of the dependence of chlorophyll a content in the phytoplankton of the Kakhovka reservoir on the combined effect of factors by means of multiple regression,” Inductive modeling of complex systems, Collected papers, Kyiv: IRTC ITS NASU, 7, pp. 147–153. (In Ukrainian)
8. Meloun, M., Militk, J., 2011. Statistical Data Analysis. A Practical Guide. Woodhead Publishing India.
   https://doi.org/10.1533/9780857097200
9. Tibshirani, R., 1996. “Regression Shrinkage and Selection via the Lasso,” Journal of the Royal Statistical Society, Series B (Methodological), 58 (1), pp. 267-288.
   https://doi.org/10.1111/j.2517-6161.1996.tb02080.x
10. Stepashko, V., 2018. “Developments and Prospects of GMDH-Based Inductive Modeling, In: Advances in Intelligent Systems and Computing II” / N. Shakhovska, V. Stepashko, Editors, AISC book series, Cham: Springer, 2018, vol. 689, pp. 474-491.
    https://doi.org/10.1007/978-3-319-70581-1_34
11. Pidnebesna, H.A., 2019. “Correlation-Based Sorting Algorithm of Inductive Modeling using Argument Rating,” Proceedings of the IEEE 14th Internnational Scientific and Technical Conference on Computer Sciences and Information Technologies (CSIT-2019), Lviv, Ukraine: LNPU, September 17-20, vol. 1, pp. 211-214.
    https://doi.org/10.1109/STC-CSIT.2019.8929761
12. Rawlings, J.O., Sastry, G.P., Dickey, D.A., 1998. Applied Regression Analysis: A Research Tool, Second Edition, Springer.
    https://doi.org/10.1007/b98890

 Received  20.03.2020