Earth Observation data for Improved Water Quality Products in freshwater reservoirs

Autor: Bresciani Mariano a, Giardino Claudia a, Maria Antonietta Dessena b, Paola Buscarinu b, Loretta Cabras b, Karin Schenk c, Thomas Hegge c, Hendrik Bernet c, Giorgos Bazdanis d, Apostolos Tzimas e
Jazyk: angličtina
Rok vydání: 2018
Předmět:
Zdroj: AIT 2018: THE IX CONFERENCE OF THE ITALIAN SOCIETY OF REMOTE SENSING, Firenze, 4-6/7-2018
info:cnr-pdr/source/autori:Bresciani Mariano a, Giardino Claudia a, Maria Antonietta Dessena b, Paola Buscarinu b, Loretta Cabras b, Karin Schenk c, Thomas Hegge c, Hendrik Bernet c, Giorgos Bazdanis d, Apostolos Tzimas e/congresso_nome:AIT 2018: THE IX CONFERENCE OF THE ITALIAN SOCIETY OF REMOTE SENSING/congresso_luogo:Firenze/congresso_data:4-6%2F7-2018/anno:2018/pagina_da:/pagina_a:/intervallo_pagine
Popis: Water resources are limited and are facing issues that are also due to climate change, which could have serious consequences for water quality. Remote sensing techniques are widely used to study the Earth's surface. They have many scientific applications, including the evaluation and quantification of changes in water quality. The SPACE-O (Space Assisted Water Quality Forecasting Platform for Optimized Decision Making in Water Supply Services) Project on the Horizon2020 program is catalyzing innovation with a service platform created to facilitate interoperability between the Earth Observation data, ecological and hydraulic models and services implemented in a Management Decision Support System (DSS). In this work we present the activities based on EO data analysis for the two inland water dams (Lake Mulargia in Sardinia-Italy and lake Aposelemis in Creta-Greek). Mulargia (principal tributary of Flumendosa) dam is located in south Sardinia island. Mulargia reservoir serves as a drinking water source for the towns of Cagliari and hinterland and 20 villages around in the greater area summing up to a population of 700.000 inhabitants. The total annual abstraction for drinking water purposes is estimated to 100 hm3. Aposelemis dam is located in south Greece at Crete island. Aposelemis reservoir serves as a drinking water source for the towns of Heraklion and Agios Nikolaos as well as local communities in the greater area summing up to a population of 300.000 inhabitants. The total annual abstraction for drinking water purposes is estimated to 17 hm3. Earth observation (EO) imagery gathered from Landsat-8 and Sentinel-2A & B are operationally used with SPACE-O to observe a set of key variables for water quality monitoring. The products, based on the physics-based Modular Inversion and Processing System (MIP) image-processing-chain, provide time-series data of the water quality parameters from 2013 to 2017. The MIP architecture systematically handles the independent properties of sensor parameters and specific optical properties as well as the radiative transfer relationships (at 1nm spectral resolution). The processing includes the acquisition or harvesting of satellite data, radiometric calibration, spatial subsetting (clipping) of region of interest, Land-Water-Cloud masking, Adjacency correction, sunglint correction coupled atmospheric and water surface correction, including aerosol and in-water Inherent Optical Properties (IOPs) retrieval, retrieval of water constituent concentrations from IOP's and Creation of quality indicators and quality control. The water quality obtained was Chlorophyll-a concentration, Turbidity, Secchi Disk Depth, Lake Surface Temperature; in addition of these parameter some ecological and warning indexes was produced (e.g Trophic Status Index, Harmful Algal Bloom Probability). Moreover, The Thermal Infra-Red Sensor (TIRS) on board of Landsat-8 data was used, in combination spectral information obtained from the optical sensor (namely, OLI), to characterise the heat fluxes and the evaporation rate in the study areas. Lake surface temperatures are retrieved from top-of-atmosphere brightness temperature in the TIRS band 10 (10.9 ?m). Then, the surface energy budget for the Mulargia and Aposelemis reservoirs was estimated using Landsat-8 derived products (i.e. surface temperature and surface albedo) and surface air temperature, wind speed, and relative humidity data from the hydro-meteorological surface meteorological stations according to evaporative fraction method. The validation of the products show a good agreement between EO vs in situ data (turbidity R2 = 0.82 and chl-a R2 = 0.72), in particular the trend of the results during the years show the same tendency. In order to be able to provide the managers of the dams a given uncertainty about the use of satellite data, as a standard part of the processing, an accuracy or quality indicator is calculated for each retrieved parameter and for each detected water pixel. This measure comprises a comprehensive range of factors that can impact the derived product quality (e.g. the estimated sun glint probability; residuals of the measured and modelled sensor radiances and subsurface reflectances; pixels affected by cloud shadow). The quality information is part of each standard geodata delivery and is visualized by different maps from which the final product masks are obtained. For Lake Mulargia 138 clouds free images produce many information about the water quality, the average chl-a was 10.8 mg/m3 (Min 0.8 and Max 77.2), and average turbidity was 5.9 ETU (Min 0.9 and Max 44). The results show that in spring and autumn period the quality of the water is lower, in particular in the period when decreasing the level of the water and when the precipitation was copious and determine a high run-off of particulate matter from the basin. For Lake Aposelemis 129 clouds free images show that the average chl-a was 16 mg/m3 (Min 1.7 and Max 45.2) with reduced spatial variability within the lake, the turbidity values show a greater inhomogeneity's with south zone of reservoir have the higher value (average 6.8 ETU; standard deviation 4.3) respect to the northern part near to the dam (average 5.2 ETU; standard deviation 3.5). The highest values were recorded in the period January-March 2015 and 2017. For both freshwater reservoirs, the trophic state judgments obtained with EO data was Mesotrophic and the evaluation of inorganic and organic fraction shows how the inorganic component was predominant in winter and the organic in summer in the Aposelemis and in spring and autumn Mulargia. In order to overcome temporal gaps, calculation of statistical means from satellite data was apply, this includes monthly and seasonal means and the minimum, maximum, quantiles (0.25,0.75), standard deviation, range and count of all images which have been taken into account for a certain pixel are extracted. The Space-O project requests additional optical very high resolution datasets from the COPERNICUS data-warehouse. Rapid-Eye images will be used to investigate the presence of cyanobacteria surface blooms, with semi-empirical RED/NIR band ratio approaches, in Mulargia Dam and floating vegetation in Aposelemis Dam with Normalized Difference Aquatic Vegetation Index algorithms. The results show a presence of patchy distribution on red cyanobacteria in 8 April and 23 October 2015 in Mulargia dam with a coverage of 2.6 and 1.4 hectares respectively and over 2.4 hectares of floating vegetation in Aposelemis Dam on 22 August 2014. All the products obtained from EO data are used in the SPACE-O modules to support the implementation of the Mulargia and Aposelemis monitoring as well as the water forecasting. SPACE-O also provides the opportunity to produce water information systems that are based on near real-time EO products.
Databáze: OpenAIRE
Externí odkaz: