Use of geospatial and multivariate statistical analysis in support of water quality monitoring of hydroelectric reservoirs
Isabel de Sousa Brandão is a PhD student in the Department of Water Resources. Her supervisors are dr.ir. C.M.M. Mannaerts and prof.dr.ing. W. Verhoef from the faculty of Geo-Information Science and Earth Observation.
Hydropower reservoirs are man-made artificial aquatic ecosystems that present a high dynamic and complexity in space and time, with interactions between its structural (dam), physical-chemical and biological components. They are important not only for their electrical power generation, but also for other functions such as water supply (e.g. irrigation, drinking and industry water), flood control, fisheries, as ecological wetland, for leisure activities and navigation. However, their construction causes diverse impacts to terrestrial and aquatic systems. In aquatic systems they interfere with the physical and chemical conditions of the water quality due to alterations of the hydrological regime of the dammed river, and with the functioning mechanisms and succession of phytoplankton communities (Tundisi et al., 2008). Terrestrial impacts include loss of fauna and flora, dislocation of population in areas which will be flooded and increase of endemic diseases. Nevertheless, these large constructions largely affect the water environment and their impoundments regularly cause changes leading to degradation of water quality. Typical effects noticed in tropical hydropower reservoirs are eutrophication that leads to recurring algae blooms which can be harmful to human health and greenhouse emissions which is a global concern among researchers around the world due to these gases potential to contribute to global warming.
Conventionally, the study of water quality within hydroelectric reservoirs is based on systematic point sampling which however does not represent the whole system, especially in the case of large lakes. Remote sensing data have a potential to observe and study these large aquatic environments because they provide synoptic information over the whole reservoir areas, by capturing the spectral signals and their variabilities that occur and reflect from the water body.
The main objective of this thesis was to integrate geospatial information with in situ water quality monitoring to improve on the cost efficiency of environmental management schemes of hydropower reservoirs in the Amazon region. The main objective was achieved as a combination of four steps. The first step consist of an investigation of phytoplankton diversity and response to environmental disturbance in a sustainable reserve located within the Tucuruí hydroelectric reservoir. Knowledge about phytoplankton community structure helps in assessing the quality of a water body. However, variables related to it are not routinely surveyed in most of the water quality monitoring programs. The research was carried out in the rainy and dry season when measurements were performed every three hours and at five different depths. A total of 40 water samples were collected to assess temporal variations of abiotic and biotic factors. Physico-chemical parameters were analysed to characterize the ecosystem and relationships between these variables and phytoplankton functional groups were statistically tested. The data were examined using analysis of variance and canonical correspondence analysis. We identified 9 functional groups in both seasons. The functional group M, which represents organisms with developed adaptations to high insolation and stable environments, had a higher relative percentage of contribution to the total biomass in the rainy season. Group P, which tends to be present in the more eutrophic lakes and is tolerant to carbon deficiency, had a higher relative percentage of contribution to the total biomass in the dry season. The results of this step indicated that the fluctuations of the water level reflected in seasonal changes of phytoplankton biomass and environmental variables. Additionally, this experiment permitted to advise on sampling strategies for monitoring phytoplankton in lakes and reservoirs.
The second step was to assess the feasibility of using medium high resolution sensors, such as Landsat-8 OLI sensor in monitoring the spatial distribution and frequency of phytoplankton blooms in the Tucuruí reservoir. Monitoring algal blooms from space is a very challenging task, which becomes particularly difficult when dealing with cyanobacteria blooms. Cyanobacteria are strategic organisms adapted to a wide variety of environmental conditions. In high concentrations, they form scum on the water surface, which is a concern for public health due to the production of toxins, as well as being a nuisance. Knowledge of the ecological role of these organisms is, therefore, essential when trying to estimate their extent from satellite-based data. In this step we present a multidisciplinary approach, based on both the ecological and the optical perspective. This approach is applied in a Brazilian Amazonian reservoir using spatial and temporal scales. We used a slope algorithm based on the red-edge bands of the OLI sensor and the slope algorithm could identify the extent of the algal bloom at both the spatial and temporal scale. Unfortunately, the performance of these algorithms is most likely affected by weather conditions and glint effects. Therefore, this study recommends that cyanobacteria or phytoplankton studies in this area ensure that their ecological functioning is carefully considered when attempting to map their occurrence using limited satellite imagery.
In step three the aim was to identify key environmental factors influencing eutrophication and associated harmful algae bloom occurrences in the Tucuruí hydropower, e.g. human influences and climate forcing (deforestation, human settlements, aquaculture, reservoir hydrological operation cycles and management, climate variations). The eutrophication of hydroelectric reservoirs is influenced by both anthropogenic and natural factors. The trophic state of a water body is an important variable when characterizing water quality, due to nutrient inputs originating from anthropogenic sources. Few studies have examined the influence of reservoir hydrodynamics on the water quality of its limnological zones. In this context, the relationships between the hydrological cycle of an Amazon reservoir and the water quality in its limnological zones with respect to factors influencing eutrophication processes were assessed herein. Data were collected on the surface area, from 2005 to 2016, at twelve stations distributed upstream the dam. Chlorophyll-a (Chl-a), water transparency, dissolved oxygen (DO), total phosphorus, orthophosphate, total suspended solids (TSS) and turbidity were determined, as they undergo alterations during the hydrological cycle and present zonal variations. Generalized linear models were used to identify the response of limnological variables in relation to the influence of the hydrological cycle on the water quality of the longitudinal zones. The results indicated that the filling and full cycles display higher eutrophication conditions than the dry and emptying cycles, with mean phosphorus values of 30.42 µgL-1 and 30.15 µgL-1. The riverine zone presented higher eutrophication conditions, with mean phosphorus values of 32.32 µgL-1, higher than the limits established in the Brazilian CONAMA 357/2005 resolution for Class 2 lentic environments (< 30µgL-1).
In the last step, the aim was to estimate the GHG emissions in the Tucuruí hydroelectric reservoir in temporal and spatial scales using geospatial analysis and to discuss if emissions are related to the eutrophication process due to anthropic activities or climate forcings. Hydroelectric power reservoirs are considered potential contributors to the greenhouse effect in the atmosphere through the emission of methane and carbon dioxide. In the last step, we combined in situ sampling and gas chromatography with geostatistical and remote sensing approaches to estimate greenhouse gas (GHG) emissions of a large hydropower reservoir. We used remote sensing data to estimate the water surface and geospatial interpolation to calculate total emissions as a function of reservoir surface area. The CH4 and CO2 gas concentrations were linearly correlated to sampling time, confirming the adequacy of the in situ sampling method to measure GHG diffusive fluxes from reservoir water surfaces. The combination of high purity (99.99%) ISO-norm gas standards with a gas chromatograph, enabled us to achieve low measurement detection limits of 0.87 ppm and 1.22 ppm, respectively, for CH4 (using a flame ionization or FID detector) and CO2 (using a thermal conductivity or TCD detector). Our results show that CO2 emissions are significantly (an order of 5.102 - 103) higher than those of CH4 in both the spatial and temporal domain for this reservoir. The total diffusive GHG emissions over a year (June 2011 to May 2012) of the Tucuruí hydropower reservoir being in operation, in units of tons of carbon, added up to 6.82 x 10³ for CH4 and 1.19 x 106 for CO2. We show that in situ GHG sampling using small floating gas chambers and high precision gas chromatography can be combined with geospatial interpolation techniques and remote sensing data to obtain estimates of diffusive GHG emissions from large water bodies with fluctuating water surfaces such as hydropower reservoirs. We recommend that more measurements and observations on these emissions are pursued in order to support and better quantify the ongoing discussions on estimates and mitigation of GHG emissions from reservoirs in the Amazon region and elsewhere in the world.
Accordingly, with results obtained in the four step investigations conducted in this research, it is possible to conclude that the synergy between geospatial analysis and in situ observation is a most desirable and efficient approach to monitor the water quality of hydroelectric reservoirs.