Thermal regime has a critical impact on the lotic environment, as maximum temperature determines the boundaries of the occurrence of aquatic species, seasonal and diurnal water temperature variations affect their bioenergetics, while the timing of specific water temperature values during the year is important in the context of spawning and migrations. However, despite the great importance of water temperature studies in the context of environmental management and fisheries, as well as the development of accurate measurement techniques, such investigations have received relatively limited attention in Poland. The current study attempted to examine the seasonal differentiation of water temperature in lowland rivers. For this purpose, water temperature was recorded from the 1st of May 2015 to the 30th of April 2019 with a temporal resolution of 30-minutes. Digital temperature reorders used to make the measurements were distributed across six sites in Jeziorka, Świder and Utrata catchments located on the Mazovian Lowland and the Southern Podlachia Lowland near Warsaw. The hydrometeorological background of the water temperature monitoring was determined on the basis of data from the Warszawa-Okęcie station and water gauging stations. On the basis of the measurement data, mean, maximum, and minimum monthly water temperatures were calculated and presented on the background of the appropriate air temperature data, while statistical distribution of the 30-minute water temperature, aggregated in a monthly timescale, was presented on the box and whiskers plots. The Ward method was used to group months similar in terms of their thermal conditions, while the Pearson correlation coefficient was applied to evaluate the strength of the relationship between water and air temperature. The results indicate that the seasonal course of water temperature follows the course of air temperature, with the highest mean monthly water temperatures recorded in July, while the lowest in January. Statistical distribution analysis of water temperature in individual months and its grouping by the Ward method allowed to identify two periods characterized by relatively stable thermal conditions and two periods of dynamic changes of water temperature. In contrast to the maximum values of water temperature, which were observed in the summer as a result of intensive solar radiation and low streamflow rates, the greatest variability of water temperature, as indicated by reference to mean daily range and standard deviation, was found in the spring months, i.e. in April and May, while the lowest in winter, from December to February. The relationship between daily mean water temperature and air temperature, established with the use of the Pearson correlation coefficient on a monthly basis, was clearly stronger during the spring increase and the autumn fall of the water temperature, which can be linked with greater vulnerability to atmospheric heat fluxes. A definitely weaker relationship was found in the winter and summer months, when greater importance can be attached to other drivers of stream temperature, like the presence of ice cover, cloudiness, riparian shading, and groundwater inflows.
The unbiased estimation of r2 between two sets of noisy neural responses