# An UAV and Satellite Multispectral Data Approach to Monitor Water Quality in Small Reservoirs

^{1}

^{2}

^{3}

^{4}

^{*}

## Abstract

**:**

## 1. Introduction

## 2. Materials and Methods

#### 2.1. Study Area

^{3}located in Spain, in the northwest (NW) of the Iberian Peninsula (Figure 1). Since 2000, the reservoir has maintained a drinking water supply service for households and facilities in the municipality. According to the data obtained during the conducted sampling, it has a maximum depth of 9.0 m. A river park was set up in 2012, when the reservoir also became navigable and a small jetty and a recreational area were installed, confirming its use as a leisure area for residents. The reservoir is located in the Atlantic Biogeographical Region, under an oceanic temperate climate, with a mean annual temperature of 19.6 °C and high precipitation (1146.2 mm) and relative humidity (90.6%) (mean annual values for five years (2013–2017) [27]). The rainiest period spans from November to February.

#### 2.2. In Situ Data

#### 2.3. Remotely Sensed Data

#### 2.3.1. Satellite Imagery Pre-Processing

#### 2.3.2. UAV Imagery Pre-Processing

#### 2.3.3. Data Analysis

- λ
_{1}: Spectral region such that R^{−1}_{λ1}is maximally sensitive to the absorption chl-a but still affected by the absorption of other constituents and backscattering: λ_{1}= 660–690 nm. - λ
_{2}: Spectral region such that R^{−1}_{λ2}is minimally sensitive to chl-a for which the absorption by other constituents is almost equal to that at λ_{1}: 710-λ_{2}-730 nm. - λ
_{3}: Spectral region minimally affected by the absorption of pigments, used to compensate for the variability in backscattering between samples: initially λ_{3}> 740 nm.

_{1}and λ

_{3}terms of the equation and gives good results when a

_{chla}(λ

_{1}) ≫ b

_{b}and a

_{chla}(λ

_{1}) ≫ a

_{tripton}(λ

_{1})+ a

_{CDOM}(λ

_{1}) [39], any a being the absorption coefficients and b

_{b}the backscattering coefficient.

_{λ3}. Band 6 Rrs data for MSI have been mostly used in 3BDA and 2BDA algorithms formulated by other authors (e.g., [46]) as it offers more reliable data for water studies inside the NIR region of the spectrum than the rest of NIR bands in MSI. Yet, this band is centered exactly in 740 nm, which is lower than the theoretical wavelength specified by [23] who established the optimal wavelength above 740 nm and obtained the best results with 740–750 nm. Therefore, just one version of these two algorithms with Rrs data from Band 7 (Figure 2, Table 2 and Table 3) was tested.

^{−1}λ

_{1}- R

^{−1}λ

_{2}) is related to the content of chl-a, but it is still affected by the variability in the backscattering of the medium; however, in our case, we also tried to apply this part of the algorithm alone with the UAV Multispectral Imagery data.

## 3. Results

#### 3.1. Water Quality

#### 3.2. Spectral Band Combinations for the Retrieval of chl-a

#### 3.2.1. Landsat 8 Imagery

^{2}being 0.858 and 0.753 for 2BDA, 0.844 and 0.712 for SABI and 0.834 and 0.6953 for NDVI models, respectively. Kab_1 was also tested as its results were considered good, but the scatterplot showed a relationship based in two defined and separated group of points.

#### 3.2.2. Sentinel 2 Imagery

^{2}being 0.718 and 0.5167 for 2BDA_2, 0.660 and 0.436 for 3BDA_2, 0.653 and 0.427 for Kab_1 and 0.659 and 0.435 for B3B2, respectively.

^{2}values but the correlation was highly dependent on one data point, so they were not considered robust. The results of the calibration/validation exercise are given (Table 8, Figure 7).

#### 3.2.3. UAV Imagery. Model Calibration

_{MOD}), for all the three sampling areas tested. (Table 9, Table 10 and Table 11, Figure 8)

^{2}(0.9816) and the highest significance level.

#### 3.3. Combined Monitoring Tool

^{−1}for blue, green, red, rededge and NIR bands, respectively (Figure 11).

## 4. Discussion

^{−1}, respectively.

^{−1}, respectively, which is similar to what was found for Sentinel 2 MSI by Warren et al. [60] for Sentinel 2 MSI. Even though other valuable comparative exercises have been done regarding the performance and associated uncertainties of the AC processor for OLI data used in our study [63], they were performed over different land covers and so were not taken as a reference for our work.

_{3}). Moreover, in agreement with our study are the results obtained by Xu et al. (2018), who also found good relationships with 3BDA, 2BDA and NDCI, although they used a different formulation of the 2BDA algorithm (B5/B4).

^{2}all above 0.96 for all bands. In field experiments, the comparison with the Pix4D processed image showed lower accuracy (overestimation) in the rededege and NIR regions, but a consistency in the spectral signature shape which reduced the uncertainty when using normalized indices.

_{1}and λ

_{2}, without removing the effect of the backscattering of the medium with the use of λ

_{3}, gave the best results with this sensor in low chl-a conditions. The optimal position of λ

_{3}in 2BDA and 3BDA algorithms were established in practice by Dall’Olmo and Gitelson [70] in 720–740 nm in turbid productive waters. The NIR band of the Rededege Micasense is centered in 840 nm, hence too far from the optimal range, and proved not useful in our study in low chl-a values. As expected by the formulation of the algorithm, when applied in turbid conditions with high chl-a during the second UAV flight, this modification did not perform well (R

^{2}= 0.4; Pearson r = 0.6) as the effect of the backscattering of a medium with a high load of suspended particles should be removed with the λ

_{3}spectral region. Therefore, in this case 2BDA and SABI were the best performing algorithm in the presence of a bloom and high chl-a concentration. Again, the NIR spectral band in Rededge Micasense, used in this version of the 2BDA model in coherence with the formulation initially used for 3BDA, is far away from the optimal 720–740 nm range established by Dall’Olmo and Gitelson [70] to be used turbid productive waters, but performed better than the version of this algorithm made with the rededge band (2BDA_2). This band is centered in 717 nm, nearer of the optimum range, but the authors describe the aim of this band λ

_{3}as to be in a spectral region where the absorption by pigments, tripton and CDOM is negligible, which is something that happens at λ > 740 [21] so the use of longer wavelengths (840 nm in our case) might be feasible. Moreover, as chl-a increases, the λ

_{3}spectral region of maximal sensitivity to chl-a shifts toward longer wavelengths, and the precise optimal position of λ

_{1}and λ

_{3}will depend on the relative importance of the interferences and on the trophic status of the waterbody [70].

## 5. Conclusions

## Author Contributions

## Funding

## Acknowledgments

## Conflicts of Interest

## References

- Ibisch, R.; Austnes, K.; Borchardt, D.; Boteler, B.; Leujak, W.; Lukat, E.; Rouillard, J.; Schmedje, U.; Lyche Solheim, A. European Assessment of Eutrophication Abatement Measures across Land-Based Sources Inland, Coastal and Marine Waters; European Environment Agency: Copenhagen, Denmark, 2016; pp. 7–9. [Google Scholar]
- Lemley, D.A.; Adams, J.B. Eutrophication. Earth Syst. Environ. Sci.
**2019**, 1, 86–90. [Google Scholar] [CrossRef] - European Environmental Agency. European Waters—Assessment of Status and Pressures; EEA Report No. 8/2012; European Environment Agency: Copenhagen, Denmark, 2012; Available online: http://www.eea.europa.eu/publications/european-waters-assessment-2012 (accessed on 12 December 2019).
- The European Parliament the Council of the European Union WFD. Directive 2000/60/EC of the European Parliament of the Council of 23 October 2000 establishing a framework for Community action in the field of water policy. OJL
**2000**, 327, 1–73. [Google Scholar] - Clean Water Act. Federal Water Pollution Control Act [As Amended Through P.L. 107–303, 27 November 2002]; United Stated Environmental Protection Agency: Washington, DC, USA, 2002. Available online: https://www.epa.gov/sites/production/files/2017-08/documents/federal-water-pollution-control-act-508full.pdf (accessed on 2 December 2019).
- Brezonik, P.; Menken, K.D.; Bauer, M. Landsat-based remote sensing of lake water quality characteristics, including chlorophyll and colored dissolved organic matter (CDOM). Lake Reserv. Manag.
**2005**, 21, 373–382. [Google Scholar] [CrossRef] - Kutser, T.; Metsamaa, L.; Strömbeck, N.; Vahtmäe, E. Monitoring cyanobacterial blooms by satellite remote sensing. Estuar. Coast. Shelf. Sci.
**2006**, 67, 303–312. [Google Scholar] [CrossRef] - Moses, W.J.; Gitelson, A.A.; Berdnikov, S.; Povazhnyy, V. Satellite estimation of chlorophyll—A concentration using the red and NIR bands of MERIS—The Azov Sea case study. IEEE Geosci. Remote Sens. Lett.
**2009**, 6, 845–849. [Google Scholar] [CrossRef] - Matthews, M.W. A current review of empirical procedures of remote sensing in inland and near-coastal transitional waters. Int. J. Remote Sens.
**2011**, 32, 6855–6899. [Google Scholar] [CrossRef] - Matthews, M.W. Eutrophication and cyanobacterial blooms in South African inland waters: 10 years of MERIS observations. Remote Sens. Environ.
**2014**, 155, 161–177. [Google Scholar] [CrossRef] - Shi, K.; Zhang, Y.; Liu, X.; Wang, M.; Qin, B. Remote sensing of diffuse attenuation coefficient of photosynthetically active radiation in Lake Taihu using MERIS data. Remote Sens. Environ.
**2014**, 140, 365–377. [Google Scholar] [CrossRef] - Giardino, C.; Bresciani, M.; Stroppiana, D.; Oggioni, A.; Morabito, G. Optical remote sensing of lakes: An overview on Lake Maggiore. J. Limnol.
**2014**, 73, 201–214. [Google Scholar] [CrossRef] [Green Version] - Dörnhöfer, K.; Oppelt, N. Remote sensing for lake research and monitoring—Recent advances. Ecol. Indic.
**2016**, 64, 105–122. [Google Scholar] [CrossRef] - Anster, A.; Alikas, K. Retrieval of Chlorophyll a from Sentinel-2 MSI Data for the European Union Water Framework Directive Reporting Purposes. Remote Sens.
**2019**, 11, 64. [Google Scholar] [CrossRef] [Green Version] - Gitelson, A. The peak near 700 nm on radiance spectra of algae and water: Relationships of its magnitude and position with chlorophyll concentration. Int. J. Remote Sens.
**1992**, 13, 3367–3373. [Google Scholar] [CrossRef] - Kirk, J.T. Light and Photosyntesis in Aquatic Ecosystems, 2nd ed.; Cambridge University Press: Cambridge, UK, 1994; p. 509. [Google Scholar]
- Mobley, C.D. Light and Water: Radiative Transfer in Natural Waters; Academic Press: Cambridge, MA, USA, 1994; p. 608. [Google Scholar]
- Morel, A.; Prieur, L. Analysis of variations in ocean color. Limnol. Oceanogr.
**1977**, 22, 709–722. [Google Scholar] [CrossRef] - Witter, D.L.; Ortiz, J.D.; Palm, S.; Heath, R.T.; Budd, J.W. Assessing the application of SeaWiFS ocean color algorithms to Lake Erie. J. Great Lakes Res.
**2009**, 35, 361–370. [Google Scholar] [CrossRef] - Horion, S.; Bergamino, N.; Stenuite, S.; Descy, J.P.; Plisnier, P.D.; Loiselle, S.A.; Cornet, Y. Optimized extraction of daily bio-optical time series derived from MODIS/aqua imagery for Lake Tanganyika, Africa. Remote Sens. Environ.
**2010**, 114, 781–791. [Google Scholar] [CrossRef] - D’all Olmo, G.; Gitelson, A.; Rundquist, D.C. Towards a unified approach for remote estimation of chlorophyll-a in both terrestrial vegetation and turbid productive waters. Geophys. Res. Lett.
**2003**, 30, 1–4. [Google Scholar] [CrossRef] [Green Version] - Gitelson, A.A.; Dall’Olmo, G.; Moses, W.; Rundquist, D.C.; Barrow, T.; Fisher, T.R.; Gurlin, D.; Holz, J. A simple semi-analytical model for remote estimation of chlorophyll-a in turbid waters: Validation. Remote Sens. Environ.
**2008**, 112, 3582–3593. [Google Scholar] [CrossRef] - Gitelson, A.A.; Gurlin, D.; Moses, W.; Barrow, T. A bio-optical algorithm for the remote estimation of the chlorophyll-a concentration in case 2 waters. Environ. Res. Lett.
**2009**, 4, 1–5. [Google Scholar] [CrossRef] - Domínguez, J.A.; Alonso, C.; Alonso, A. Remote sensing as a tool for monitoring water quality parameters for Mediterranean Lakes of European Union water framework directive (WFD) and as a system of surveillance of cyanobacterial harmful algae blooms (SCyanoHABs). Environ. Monit. Assess.
**2011**, 181, 317–334. [Google Scholar] [CrossRef] - Li., J.; Roy, D.P. A Global Analysis of Sentinel-2A, Sentinel-2B and Landsat-8 Data Revisit Intervals and Implications for Terrestrial Monitoring. Remote Sens.
**2017**, 9, 902. [Google Scholar] [CrossRef] [Green Version] - Pahlevan, N.; Chittimalli, S.K.; Balasubramanian, S.V.; Vellucci, V. Sentinel-2/Landsat-8 product consistency and implications for monitoring aquatic systems. Remote Sens. Environ.
**2019**, 220, 19–29. [Google Scholar] [CrossRef] - Meteogalicia. Available online: https://www.meteogalicia.gal/ (accessed on 2 February 2018).
- GAIA. Available online: https://gaia.xunta.es/plataforma/temas/agua/roaga/seguimiento-embalses/consulta-mediciones (accessed on 5 January 2020).
- Welschmeyer, N. Fluorometric analysis of chlorophyll a in the presence of chlorophyll b and pheopigments. Limnol. Oceanogr.
**1994**, 39, 1985–1992. [Google Scholar] [CrossRef] - Gordon, H.R.; McCunley, W.R. Estimation of the depth of sunlight penetration in the sea for remote- sensing. Appl. Opt.
**1975**, 4, 413–416. [Google Scholar] [CrossRef] [PubMed] - Jeffrey, S.W.; Humphrey, G.F. New spectrophotometric equations for determining chlorophylls a, b, c1 and c2 in higher plants, algae and natural phytoplancton. Biochem. Physiol. Pflanzen.
**1975**, 167, 191–194. [Google Scholar] [CrossRef] - American Public Health Association, American Water Works Association and Water Environmental Federation. APHA Standard Methods for the Examination of Water and Wastewater, 20th ed.; American Public Health Association, American Water Works Association and Water Environmental Federation: Washington, DC, USA, 1998; p. 874.
- Quesada, A. Ficobilinas o Ficobiliproteínas. In Estudio de las Aguas Continentales Mediante Teledetección; Domínguez, J.A., Marcos, C., Chao, Y., Delgado, G., Rodríguez, D., Eds.; Universidad Nacional de Educación a Distancia: Madrid, Spain, 2011; pp. 82–85. [Google Scholar]
- De Keukelaere, L.; Sterckx, S.; Adriaensen, S.; Knaeps, E.; Reusen, I.; Giardino, C.; Bresciani, M.; Hunter, P.; Neil, C.; Van der Zande, D.; et al. Atmospheric correction of Landsat-8/OLI and Sentinel-2/MSI data using iCOR algorithm: Validation for coastal and inland Waters. Eur. J. Remote Sens.
**2018**, 51, 525–542. [Google Scholar] [CrossRef] [Green Version] - U.S. Geological Survey. LANDSAT 8 Surface Reflectance Code (Lasrc) Product Guide. V 2.0; U.S. Geological Survey: Reston, VA, USA, 2019; p. 39.
- Pahlevan, N.; Schott, J.R.; Franz, B.A.; Zibordi, G.; Markham, B.; Bailey, S.; Schaaf, C.B.; Ondrusek, M.; Greb, S.; Strait, C.M. Landsat 8 remote sensing reflectance (Rrs) products: Evaluations, intercomparisons, and enhancements. Remote Sens. Environ.
**2017**, 190, 289–301. [Google Scholar] [CrossRef] - Bernardo, N.; Watanabe, F.; Rodrigues, T.; Alcantara, E. Atmospheric correction issues for retrieving total suspended matter concentrations in inland waters using OLI/Landsat-8 image. Adv. Space Res.
**2017**, 59, 2335–2348. [Google Scholar] [CrossRef] [Green Version] - Watanabe, F.; Alcantara, E.; Rodrigues, T.; Rotta, L.; Bernardo, N.; Imai, N. Remote sensing of the chlorophyll-a based on OLI/Landsat-8 and MSI/Sentinel-2A (Barra Bonita reservoir, Brazil). Ann. Braz. Acad. Sci.
**2017**, 90, 1–14. [Google Scholar] [CrossRef] [Green Version] - R Core Team. R: A Language and Environment for Statistical Computing; R Foundation for Statistical Computing: Vienna, Austria, 2018; Available online: https://www.R-project.org/ (accessed on 5 May 2020).
- Bivand, R.; Keitt, T.; Rowlingson, B. rgdal: Bindings for the ‘Geospatial’ Data Abstraction Library. R Package Version 1.4-4. 2019. Available online: https://CRAN.R-project.org/package=rgdal (accessed on 5 May 2020).
- Hijmans, R.J. raster: Geographic Data Analysis and Modeling. R Package Version 2.9-5. 2019. Available online: https://CRAN.R-project.org/package=raster (accessed on 5 May 2020).
- Wickham, H.; François, R.; Henry, L.; Müller, K. dplyr: A Grammar of Data Manipulation. R Package. Version 0.8.1. 2019. Available online: https://CRAN.R-project.org/package=dplyr (accessed on 5 May 2020).
- Matt Dowle and Arun Srinivasan (2019). data.table: Extension of ‘data.frame’. R Package Version 1.12.2. 2019. Available online: https://CRAN.R-project.org/package=data.table (accessed on 5 May 2020).
- Gitelson, A.A.; Gritz, Y.; Merzlyak, M.N. Relationship between leaf chlorophyll content and spectral reflectance and algorithms for non-destructive chlorophyll assessment in higher plant leaves. J. Plant Physiol.
**2003**, 160, 271–282. [Google Scholar] [CrossRef] - Dall’Olmo, G.; Gitelson, A.A. Effect of bio-optical parameter variability on the remote estimation of chlorophyll-a concentration in turbid productive waters: Experimental results. Appl. Opt.
**2005**, 44, 412–422. [Google Scholar] [CrossRef] [Green Version] - Xu, M.; Hongxing, L.; Beck, R.; Lekki, J.; Yang, B.; Shu, S.; Kang, E.L.; Anderson, R.; Johansen, R.; Emery, E.; et al. A spectral space partition guided ensemble method for retrieving chlorophyll-a concentration in inland waters from Sentinel-2A satellite imagery. J. Great Lakes Res.
**2019**, 45, 454–465. [Google Scholar] [CrossRef] - Kabbara, N.; Benkhelil, J.; Awad, M.; Barale, V. Monitoring water quality in the coastal area of Tripoli (Lebanon) using High-resolution satellite data. ISPRS J. Photogramm. Remote Sens.
**2008**, 63, 488–495. [Google Scholar] [CrossRef] - Alawadi, F. Detection of surface algal blooms using the newly developed algorithm surface algal bloom index (SABI). Proc. Int. Soc. Opt. Eng.
**2010**, 7825, 1–14. [Google Scholar] [CrossRef] [Green Version] - Brivio, P.A.; Giardino, C.; Zilioli, E. Determination of chlorophyll concentration changes in Lake Garda using an image-based radiative transfer code for Landsat TM images. Int. J. Remote Sens.
**2001**, 22, 487–502. [Google Scholar] [CrossRef] - Mishra, S.; Mishra, D.R. Normalized difference chlorophyll index: A novel model for remote estimation of chlorophyll-a concentration in turbid productive waters. Remote Sens. Environ.
**2012**, 117, 394–406. [Google Scholar] [CrossRef] - Boucher, J.; Weathers, K.C.; Norouzi, H.; Steele, B. Assessing the effectiveness of Landsat 8 chlorophyll-a retrieval algorithms for regional freshwater monitoring. Ecol. Appl.
**2018**, 28, 1044–1054. [Google Scholar] [CrossRef] [Green Version] - Walsby, A.E.; Hayes, P.K.; Boje, R.; Stal, L.J. The selective advantage of buoyancy provided by gas vesicles for planktonic cyanobacteria in the Baltic Sea. New Phytol.
**1997**, 136, 407–417. [Google Scholar] [CrossRef] - Agha, R.; Cires, S.; Wörmer, L.; Domínguez, J.A.; Quesada, A. Multi-scale strategies for the monitoring of fresh- water cyanobacteria: Reducing the sources of uncertainty. Water Res.
**2012**, 46, 3043–3053. [Google Scholar] [CrossRef] - Bresciani, M.; Cazzaniga, I.; Austoni, M.; Sforzi, T.; Buzzi, F.; Morabito, G.; Giardino, C. Mapping phytoplankton blooms in deep subalpine lakes from Sentinel-2A and Landsat-8 M. Hydrobiologia
**2018**, 824, 197–214. [Google Scholar] [CrossRef] [Green Version] - Woodcock, C.E. Uncertainty in Remote Sensing. In Uncertainty in Remote Sensing and GIS; Foody, G.M., Atkinson, P.A., Eds.; John Wiley & Sons, Ltd.: Hoboken, NJ, USA, 2002; pp. 19–24. [Google Scholar] [CrossRef]
- Schott, J.R. Remote Sensing: The Image Chain Approach; Oxford University Press: New York, NY, USA, 1997. [Google Scholar] [CrossRef]
- Loew, A.; Bell, W.; Brocca, L.; Bulgin, C.E.; Burdanowitz, J.; Calbet, X.; Donner, R.V.; Ghent, D.; Gruber, A.; Kaminski, T.; et al. Validation practices for satellite-based Earth observation data across communities. Rev. Geophys.
**2017**, 55, 779–817. [Google Scholar] [CrossRef] [Green Version] - Yang, Z.; Anderson, Y. Estimating Chlorophyll-A Concentration in a Freshwater Lake Using Landsat 8 Imagery. J. Environ. Earth Sci.
**2016**, 6, 134–142. [Google Scholar] - Toming, K.; Kutser, T.; Laas, A.; Sepp, M.; Paavel, B.; Nõges, T. First experiences in mapping lake water quality parameters with sentinel-2 MSI imagery. Remote Sens.
**2016**, 8, 640. [Google Scholar] [CrossRef] [Green Version] - Warren, M.A.; Simis, S.G.H.; Martinez-Vicente, V.; Poser, K.; Bresciani, M.; Alikas, K.; Spyrakos, E.; Giardino, C.; Ansper, A. Assessment of atmospheric correction algorithms for the Sentinel-2A MultiSpectral Imager over coastal and inland waters. Remote Sens. Environ.
**2019**, 225, 267–289. [Google Scholar] [CrossRef] - Pereira-Sandoval, M.; Urrego, P.; Ruiz-Verdú, A.; Tenjo, C.; Delegido, J.; Soria-Perpinyà, X.; Vicente, E.; Soria, J.; Moreno, J. Calibration and validation of algorithms for the estimation of chlorophyll-a concentration and Secchi depth in inland waters with Sentinel-2. Limnetica
**2019**, 38, 471–487. [Google Scholar] [CrossRef] - Ilory, C.O.; Pahlevan, N.; Knudby, A. Analyzing Performances of Different Atmospheric Correction Techniques for Landsat 8: Application for Coastal Remote Sensing. Remote Sens.
**2019**, 11, 469. [Google Scholar] [CrossRef] [Green Version] - Doxani, G.; Vermote, E.; Roger, J.C.; Gascon, F.; Adriaensen, S.; Frantz, D.; Hagolle, O.; Hollstein, A.; Kirches, G.; Li, F.; et al. Atmospheric correction inter-comparison exercise. Remote Sens.
**2018**, 10, 352. [Google Scholar] [CrossRef] [Green Version] - Moses, W.; Sterckx, S.; Montes, M.J.; De Keukelaere, L.; Knaeps, E. Atmospheric Correction for Inland Waters. In Bio-Optical Modeling and Remote Sensing of Inland Waters; Mishra, D.R., Ogashawara, I., Gitelson, A.A., Eds.; Elsevier Inc.: Amsterdam, The Netherlands, 2017; pp. 69–100. [Google Scholar] [CrossRef]
- Mishra, D.R.; Mishra, S. Plume and bloom: Effect of the Mississippi River diversion on the water quality of Lake Pontchartrain. Geocarto Int.
**2010**, 25, 555–568. [Google Scholar] [CrossRef] - Wang, D.; Ma, R.; Xue, K.; Loiselle, S.A. The Assessment of Landsat-8 OLI Atmospheric Correction Algorithms for Inland Waters. Remote Sens.
**2019**, 11, 169. [Google Scholar] [CrossRef] [Green Version] - Ha, N.T.T.; Thao, N.T.P.; Koike, K.; Nhuan, M.T. Selecting the Best Band Ratio to Estimate Chlorophyll-a Concentration in a Tropical Freshwater Lake Using Sentinel 2A Images from a Case Study of Lake Ba Be (Northern Vietnam). Int. J. Geo-Inf.
**2017**, 6, 290. [Google Scholar] [CrossRef] - Arango, J.G.; Nairn, R.W. Prediction of Optical and Non-Optical Water Quality Parameters in Oligotrophic and Eutrophic Aquatic Systems Using a Small Unmanned Aerial System. Drones
**2020**, 4, 1. [Google Scholar] [CrossRef] [Green Version] - Cao, S.; Danielson, B.; Clare, S.; Koenig, S.; Campos-Vargas, C.; Sánchez-Azofeifa, A. Radiometric calibration assessments for UAS-borne multispectral cameras: Laboratory and field protocols. J. Photogramm. Remote Sens.
**2019**, 149, 142–145. [Google Scholar] [CrossRef] - Dall’ Olmo, G.; Gitelson, A. Effect of Bio-Optical Parameter Variability and Concentration in Turbid Productive Waters: Remote Estimation of Chlorophyll-a Uncertainties in Reflectance Measurements on the Modeling Results. Appl. Opt.
**2006**, 45, 3577–3592. Available online: http://digitalcommons.unl.edu/natrespapers/259 (accessed on 18 June 2019). - Beck, R.; Zhan, S.; Liu, H.; Tong, S.; Yang, B.; Xu, M. Comparison of satellite reflectance algorithms for estimating chlorophyll-a in a temperate reservoir using coincident hyperspectral aircraft imagery and dense coincident surface observations. Remote Sens. Environ.
**2016**, 178, 15–30. [Google Scholar] [CrossRef] [Green Version] - Bricaud, A.; Babin, M.; Morel, A.; Claustre, H. Variability in the chlorophyll-specific absorption coefficients of natural phytoplankton: Analysis and parameterization. J. Geophys. Res.
**1995**, 100, 321–332. [Google Scholar] [CrossRef] - Tyler, A.; Hunter, P.D.; Spyrakos, E.; Groomb, E.; Constantinescu, A.M.; Kitchen, J. Developments in Earth observation for the assessment and monitoring of inland, transitional, coastal and shelf-sea waters. Sci. Total Environ.
**2016**, 572, 1307–1321. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Stumpf, R.P.; Tyler, M.A. Satellite detection of bloom and pigment distributions in estuaries. Remote Sens. Environ.
**1988**, 24, 385–404. [Google Scholar] [CrossRef]

**Figure 1.**Localization of the study site in Europe (

**A**), in the NW of the Iberian Peninsula (

**B**) and aerial picture of the reservoir with in situ sampling points both from the periodic monitoring [28] and from the two Unmanned Aerial Vehicle (UAV) flights (same sampling points for 2017 and 2018) (

**C**).

**Figure 2.**Graphical representation of spatial resolution and bandwidths of the three multispectral sensors used. Notation: S (Sentinel 2, MSI), L (Landsat 8, OLI) and R (Rededge Micasense), followed by the band number in each sensor.

**Figure 4.**Best performing models (models using the SBC: (

**a**) 2BDA; (

**b**) NDVI; (

**c**) SABI and (

**d**) Kab_1) for retrieving chl-a from Landsat 8 OLI data. The code color of the points reflects the time difference between the in situ data and the satellite image acquisition data: white: 0 days; light grey: 1 day; dark grey: 2 days and black: 3 days.

**Figure 5.**Validation scatter plots for the best performing models for retrieving chl-a using Landsat 8 OLI data. Models using the SBC: (

**a**) 2BDA; (

**b**) NDVI; (

**c**) SABI and (

**d**) Kab_1). Line 1:1 is also shown.

**Figure 6.**Best performing models for retrieving chl-a from Sentinel 2 MSI data (models using the SBC: (

**a**) 2BDA_2; (

**b**) 3BDA_2; (

**c**) Kab_1 and (

**d**) B3B2). The color code of the points reflects the time difference between the in situ data and the satellite image acquisition data: white: 0 day; dark grey: 2 days and black: 3 days (there are no match-ups with 1 day difference).

**Figure 7.**Validation scatter plots for the best performing models for retrieving chl-a from Sentinel 2 MSI data. Models using the SBC: (

**a**) 2BDA_2; (

**b**) 3BDA_2; (

**c**) Kab_1 and (

**d**) B3B2. Line 1:1 is also shown.

**Figure 8.**Best performing models for retrieving chl-a from Rededge Micasense on board UAV. 3BDA_MOD in (

**a**) a low-chl-a condition which corresponds to 2017 flight data, and (

**b**) and (

**c**) high chl-a condition, which corresponds with 2018 flight data. The results correspond to the calculations made with data included in a 0.5, 1.0 and 1.5-m. buffers, which are shown following this order from top to base.

**Figure 9.**Best performing models for retrieving chl-a from Rededge Micasense on board UAV when the data of both flights are combined. Models applying Spectral Band Combinations (

**a**) B3B1; (

**b**) GB1 and (

**c**) G/B are shown.

**Figure 10.**Application of the B3B1 algorithm to the images corresponding to the 2017 flight (low-chl-a) (

**a**) and 2018 flight (high chl-a) (

**b**). Legend indicates µgr/L of chl-a.

**Figure 11.**Box-whiskers plot comparing the spectral signature of Sentinel 2 MSI and Rededge Micasense sensors for the images acquired on 10/02/2018. Figure (

**a**) shows the results for the outer pixels in the reservoir. Figure (

**b**) shows the results for the central pixels in the reservoir, and Figure (

**c**) shows the results for the entire reservoir.

**Figure 12.**Graphical results of the combining monitoring approach for chl-a in September and October 2017 and 2018. Thematic chl-a (µg/l) maps for Landsat 8 OLI, Sentinel 2 (A- B) MSI and UAV Rededge Micasense flights. The date column shows the satellite overpass. The in situ column shows the results of the periodic monitoring in the reservoir in the nearest date available (BC and BP sampling points in Figure 1) in the same color code (bottom color bar).

**Table 1.**Summary statistics of the chl-a values (μgr/L) registered in the two sampling points of the reservoir during the periodic monitoring during the 3 years of the study period. (n is the total number of samples available for each year and sample point BC and BP). Source: [28].

BC | BP | |||||||||
---|---|---|---|---|---|---|---|---|---|---|

Mean | Min | Max | SD | n | Mean | Min | Max | SD | n | |

2016 | 26.01 | 6.29 | 134.34 | 27.14 | 22 | 314.49 | 2.42 | 1779.57 | 574.77 | 12 |

2017 | 11.11 | 3.09 | 21.16 | 4.76 | 45 | 20.96 | 2.86 | 89.70 | 18.75 | 43 |

2018 | 64.47 | 2.40 | 1028.76 | 216.23 | 23 | 130.31 | 2.65 | 1353.66 | 339.85 | 22 |

**Table 2.**Bandwidths, central wavelengths (nm) and spatial resolution (m) of the evaluated sensors bands. In case of MSI Sentinel 2A data are given.

L8 OLI. | B1 | B2 | B3 | B4 | B5 | ||||

Bandwidth | 20 | 65 | 75 | 50 | 40 | ||||

Central wavelength | 443 | 482 | 562 | 655 | 865 | ||||

Resolution. | 30 | 30 | 30 | 30 | 30 | ||||

S2A MSI | B1 | B2 | B3 | B4 | B5 | B6 | B7 | B8 | B8a |

Bandwidth | 21 | 66 | 36 | 31 | 15 | 15 | 20 | 106 | 21 |

Central wavelength | 443 | 492 | 560 | 665 | 704 | 740 | 783 | 833 | 865 |

Resolution. | 60 | 10 | 10 | 10 | 20 | 20 | 20 | 10 | 20 |

Rededge Mic. | B1 | B2 | B3 | B5 | B4 | ||||

Bandwidth | 20 | 20 | 10 | 10 | 40 | ||||

Central wavelength | 475 | 560 | 668 | 717 | 840 | ||||

Resolution. | 0.08 |

**Table 3.**SBC adapted to the OLI, MSI and Rededge sensors on board Landsat 8, Sentinel 2 and UAV platforms.

Algorithm | Band Math L8 OLI | Band Math S2 MSI | Band Math Rededge | Reference |
---|---|---|---|---|

Kab 1 (Rrs) | 1.67−3.94*ln(B2) +3.78*ln(B3) | 1.67−3.94*ln(B2) +3.78*ln(B3) | 1.67−3.94*ln(B1) +3.78*ln(B2) | [47] |

Kab 2 (Rrs) | 6.92274−5.7581*(ln(B1)/ln(B3) | [47] | ||

SABI | (B5−B4)/(B2+B3) | (B8A−B4)/(B2+B3) | (B4−B3)/(B1+B2) | [48] |

KIVU | (B2−B4)/B3 | (B2−B4)/B3 | (B1−B3)/B2 | [49] |

NDCI | (B5−B4)/(B5+B4) | (B5−B3)/(B5+B3) | [50] | |

NDVI | (B5−B4)/(B5+B4) | (B4−B3)/(B4+B3) | [50] | |

2BDA_1 | B5/B4 | B6/B4 | B4/B3 | [22] |

2BDA_2 | B7/B4 | B5/B3 | [22] | |

3BDA_1 | (B4^{−1} − B5^{−1}) * B6 | (B3^{−1} − B5^{−1}) * B4 | [21] | |

3BDA_2 | (B4^{−1}−B5^{−1})*B7 | [21] | ||

3BDA_MOD | (B3^{−1}−B5^{−1}) | |||

B3B1 | (B3−B1)/(B3+B1) | (B2−B1)/(B2+B1) | Normalized indices | |

B3B2 | (B3−B2)/(B3+B2) | (B3−B2)/(B3+B2) | ||

GB1 | B3/B1 | B2/B1 | Simple ratio | |

GB2 | B3/B2 | B3/B2 | Simple ratio | |

GR | B3/B4 | B3/B4 | B2/B3 | Simple ratio |

**Table 4.**Summary of the water analysis results of the in situ samplings synchronous with the UAV flights (DOC: Dissolved Organic Carbon; TSS: Total Suspended Solids; EC: Electrical Conductivity, P total: Total Phosphorous).

2017 | 2018 | |||||
---|---|---|---|---|---|---|

Mean | Max | Min | Mean | Max | Min | |

Chlorophyll a (µgr/L) | 2.19 | 2.68 | 1.34 | 93.04 | 99.3 | 89.84 |

Phycocyanin (µgr/L) | 0.18 | 0.24 | 0.13 | 19.03 | 27.21 | 18.86 |

Turbidity (NTU) | 4.16 | 5.20 | 3.07 | 2.3 | 2.8 | 1.4 |

Sechi Disc Depth (m.) | 1.7 | 2.0 | 1.5 | 1.6 | 1.75 | 1.20 |

pH Surface | 7.18 | 7.26 | 7.04 | 7.22 | 7.32 | 7.08 |

DOC (mg/L) | 2.22 | 2.10 | 2.40 | 2.77 | 2.99 | 2.57 |

TSS (mg/L) | 3.20 | 6.8 | 1.2 | 21.5 | 27.2 | 18.0 |

OD sup (mg/L) | 9.72 | 9.90 | 9.47 | 10.5 | 10.71 | 10.36 |

Temp surface (°C) | 16.83 | 16.87 | 16.76 | 16.33 | 16.43 | 16.15 |

EC surface (µS/cm) | 129.6 | 131.0 | 128.0 | 127.0 | 127.0 | 127.0 |

P total (mg/L) | 0.021 | 0.013 | ||||

Ammonium (mg/L) | < 0.05 | |||||

Nitrite (mg/L) | 0.057 | |||||

Nitrate (mg/L) | 9.27 | |||||

N total (mg/L) | 2.20 | 2.76 |

**Table 5.**Linear model coefficients and linear fit parameters for all the Spectral Band Combinations (SBC) tested for the retrieval of chl-a with multispectral Landsat 8 OLI data. The models were done with Ln (Chl-a). n = 13.

SBC | Intercept (a) | Slope (b) | R^{2} | Pearson r | p Value | Sig. Code ^{1} |
---|---|---|---|---|---|---|

SABI | 1.979 | 0.575 | 0.750 | 0.866 | 0.0001 | **** |

KIVU | 2.084 | −1.127 | 0.504 | −0.710 | 0.0066 | *** |

NDVI | 1.923 | 1.448 | 0.749 | 0.865 | 0.0001 | *** |

2BDA | 1.619 | 0.350 | 0.764 | 0.874 | 0.0001 | **** |

Kab 1 | 1.534 | 0.177 | 0.636 | 0.797 | 0.0011 | *** |

Kab 2 | 2.797 | −0.590 | 0.403 | −0.635 | 0.0198 | ** |

B3B1 | 2.077 | 0.961 | 0.479 | 0.692 | 0.0087 | *** |

B3B2 | 2.020 | 1.654 | 0.615 | 0.784 | 0.0015 | *** |

GB1 | 2.083 | 0.122 | 0.233 | 0.483 | 0.0945 | * |

GB2 | 1.521 | 0.517 | 0.577 | 0.759 | 0.0026 | ** |

GR | 1.622 | 0.681 | 0.082 | 0.286 | 0.3431 |

^{1}(*) p < 0.1; (**) p < 0.05; (***) p < 0.01; (****) p < 0.001.

RMSE | NRMSE | MAPE | Bias | |
---|---|---|---|---|

LANDSAT 8 OLI | ||||

SABI | 6.30 | 39.28 | 69.90 | 0.94 |

2BDA | 6.46 | 40.25 | 65.69 | 1.07 |

NDVI | 6.26 | 39.05 | 65.83 | 1.13 |

Kab_1 | 5.47 | 34.99 | 57.26 | −0.02 |

**Table 7.**Linear model coefficients and linear fit parameters for all the Spectral Band Combinations tested for the retrieval of chl-a with multispectral Sentinel 2 MSI data. The models were done with in situ chl-a. n = 23.

Index | Intercept (a) | Slope (b) | R^{2} | Pearson r | p Value | Sig. Code ^{1} |
---|---|---|---|---|---|---|

SABI | 5.885 | 31.716 | 0.112 | 0.335 | 0.1274 | |

KIVU | 24.360 | −28.930 | 0.296 | −0.544 | 0.0089 | *** |

NDCI | 10.130 | 61.300 | 0.213 | 0.461 | 0.0306 | ** |

2BDA_1 | 29.516 | −9.168 | 0.174 | −0.417 | 0.0532 | * |

2BDA_2 | −30.740 | 22.850 | 0.702 | 0.837 | 0.0000 | **** |

3BDA_1 | 14.515 | 3.597 | 0.006 | 0.081 | 0.7203 | |

3BDA_2 | 7.134 | 26.795 | 0.532 | 0.729 | 0.0001 | **** |

Kab_1 | −5.064 | 6.361 | 0.679 | 0.824 | 0.0000 | **** |

B3B2 | 8.662 | 61.763 | 0.673 | 0.820 | 0.0000 | **** |

GB2 | 1.611 | 10.119 | 0.656 | 0.810 | 0.0000 | **** |

GR | 2.766 | 6.108 | 0.073 | 0.270 | 0.2234 |

^{1}(*) p < 0.1; (**) p < 0.05; (***) p < 0.01; (****) p < 0.001.

RMSE | NRMSE | MAPE | Bias | |
---|---|---|---|---|

SENTINEL 2 MSI | ||||

3BDA_2 | 4.44 | 14.67 | 41.17 | 0.25 |

2BDA_2 | 7.94 | 26.24 | 87.18 | −0.94 |

B3B2 | 10.22 | 33.76 | 106.33 | 3.44 |

Kab_1 | 9.36 | 30.93 | 99.81 | 2.71 |

GB2 | 8.94 | 29.55 | 93.80 | 2.11 |

**Table 9.**Linear model coefficients and linear fit parameters for all the Spectral Band Combinations tested for the retrieval of chl-a with multispectral data from Rededge Micasense on board the UAV platform for the 2017 flight. The models were done with in situ chl-a and the median of the reflectance values included in a buffer of 0.5 m. around the sampling point.

Index | Interc. (a) | Slope (b) | R^{2} | Pearson r | p Value | Sig. Code ^{1} |
---|---|---|---|---|---|---|

SABI | 2.073 | 3.085 | 0.264 | 0.514 | 0.3755 | |

KIVU | 3.665 | −9.405 | 0.711 | −0.843 | 0.0727 | * |

NDCI | 4.091 | 28.404 | 0.603 | 0.776 | 0.1225 | |

NDVI | 2.088 | 2.778 | 0.363 | 0.603 | 0.2819 | |

2BDA | 0.970 | 1.109 | 0.267 | 0.517 | 0.3721 | |

3BDA | 3.262 | 7.039 | 0.046 | 0.215 | 0.7279 | |

3BDA_MOD | 3.715 | 0.2615 | 0.945 | 0.972 | 0.0055 | *** |

2BDA_2 | −12.031 | 16.256 | 0.597 | 0.773 | 0.1256 | |

B3B1 | 1.371 | 9.470 | 0.037 | 0.193 | 0.7549 | |

GB1 | −2.578 | 4.008 | 0.038 | 0.195 | 0.7526 |

^{1}(*) p < 0.1(**) p < 0.05; (***) p < 0.01; (****) p < 0.001.

**Table 10.**Linear model coefficients and linear fit parameters for all the Spectral Band Combinations tested for the retrieval of chl-a with multispectral data from Rededge Micasense on board the UAV platform for the 2017 flight. The models were done with in situ chl-a and the median of the reflectance values included in a buffer of 1.0 m. around the sampling point.

Index | Interc (a) | Slope (b) | R^{2} | Pearson r | p Value | Sig. Code ^{1} |
---|---|---|---|---|---|---|

SABI | 2.083 | 3.132 | 0.266 | 0.515 | 0.3738 | |

KIVU | 3.708 | −9.950 | 0.733 | −0.856 | 0.0638 | * |

NDCI | 4.124 | 28.100 | 0.527 | 0.726 | 0.1647 | |

NDVI | 2.100 | −2.823 | 0.369 | −0.607 | 0.2773 | |

2BDA | 0.940 | 1.1485 | 0.274 | 0.524 | 0.3648 | |

3BDA | 2.387 | 1.273 | 0.001 | 0.037 | 0.9521 | |

3BDA_MOD | 3.830 | 0.276 | 0.927 | 0.963 | 0.0085 | *** |

2BDA_2 | −11.787 | 16.039 | 0.520 | 0.721 | 0.1688 | |

B3B1 | 1.599 | 6.787 | 0.020 | 0.143 | 0.8186 | |

GB1 | −1.242 | 2.881 | 0.021 | 0.145 | 0.8158 |

^{1}(*) p < 0.1; (**) p < 0.05; (***) p < 0.01; (****) p < 0.001.

**Table 11.**Linear model coefficients and linear fit parameters for all the Spectral Band Combinations tested for the retrieval of chl-a with multispectral data from Rededge Micasense on board the UAV platform for the 2017 flight. The models were done with in situ chl-a and the median of the reflectance values included in a buffer of 1.5 m. around the sampling point.

Index | Interc. (a) | Slope (b) | R^{2} | Pearson r | p Value | Sig. Code ^{1} |
---|---|---|---|---|---|---|

SABI | 2.078 | 3.499 | 0.300 | 0.547 | 0.3395 | |

KIVU | 3.696 | −9.954 | 0.719 | −0.848 | 0.0695 | * |

NDCI | 4.217 | 28.646 | 0.601 | 0.775 | 0.1236 | |

NDVI | 2.099 | −3.047 | 0.401 | −0.633 | 0.2514 | |

2BDA | 0.791 | 1.292 | 0.311 | 0.558 | 0.3285 | |

3BDA | 2.743 | 3.462 | 0.017 | 0.130 | 0.8344 | |

3BDA_MOD | 3.803 | 0.264 | 0.946 | 0.972 | 0.0054 | *** |

2BDA_2 | −12.06 | 16.41 | 0.594 | 0.771 | 0.1272 | |

B3B1 | 1.729 | 5.301 | 0.8504 | 0.117 | 0.8504 | |

GB1 | −0.514 | 2.271 | 0.014 | 0.120 | 0.8471 |

^{1}(*) p < 0.1; (**) p < 0.05; (***) p < 0.01; (****) p < 0.001.

**Table 12.**Linear model coefficients and linear fit parameters for all the Spectral Band Combinations tested for the retrieval of chl-a with multispectral data from Rededge Micasense on board the UAV platform for the 2018 flight. The models were done with in situ Chl-a and the median of the reflectance values included in a buffer of 0.5 m. around the sampling point.

Index | Interc. (a) | Slope (b) | R^{2} | Pearson r | p Value | Sig. Code ^{1} |
---|---|---|---|---|---|---|

SABI | 90.672 | 28.895 | 0.912 | 0.955 | 0.0449 | ** |

KIVU | 102.65 | −80.314 | 0.912 | −0.955 | 0.0445 | ** |

NDCI | 97.157 | 31.436 | 0.320 | 0.566 | 0.4335 | |

NDVI | 91.359 | 20.330 | 0.753 | 0.868 | 0.1319 | |

2BDA | 81.783 | 9.017 | 0.867 | 0.932 | 0.0684 | * |

3BDA | 94.014 | 2.844 | 0.012 | 0.112 | 0.8875 | |

3BDA_MOD | 96.948 | 0.130 | 0.433 | 0.658 | 0.3413 | |

2BDA_2 | 78.540 | 18.710 | 0.290 | 0.538 | 0.4613 | |

B3B1 | 64.07 | 112.43 | 0.2172 | 0.466 | 0.534 | |

GB1 | 41.240 | 30.550 | 0.210 | 0.459 | 0.5408 |

^{1}(*) p < 0.1; (**) p < 0.05; (***) p < 0.01; (****) p < 0.001.

**Table 13.**Linear model coefficients and linear fit parameters for all the Spectral Band Combinations tested for the retrieval of chl-a with multispectral data from Rededge Micasense on board the UAV platform for the 2018 flight. The models were done with in situ Chl-a and the median of the reflectance values included in a buffer of 1.0 m. around the sampling point.

Index | Interc. (a) | Slope (b) | R^{2} | Pearson r | p Value | Sig. Code ^{1} |
---|---|---|---|---|---|---|

SABI | 90.684 | 29.020 | 0.911 | 0.9542 | 0.0457 | ** |

KIVU | 102.671 | −80.930 | 0.907 | −0.952 | 0.0477 | ** |

NDCI | 96.906 | 29.599 | 0.288 | 0.536 | 0.4634 | |

NDVI | 91.375 | 20.373 | 0.752 | 0.867 | 0.1330 | |

2BDA | 81.747 | 9.068 | 0.865 | 0.930 | 0.0696 | * |

3BDA | 93.583 | 1.587 | 0.004 | 0.064 | 0.9358 | |

3BDA_MOD | 93.837 | 0.127 | 0.410 | 0.640 | 0.3591 | |

2BDA_2 | 79.550 | 17.390 | 0.256 | 0.506 | 0.4937 | |

B3B1 | 68.00 | 97.26 | 0.1573 | 0.396 | 0.6033 | |

GB1 | 48.590 | 26.230 | 0.151 | 0.388 | 0.6114 |

^{1}(*) p < 0.1; (**) p < 0.05; (***) p < 0.01; (****) p < 0.001.

**Table 14.**Linear model coefficients and linear fit parameters for all the Spectral Band Combinations tested for the retrieval of chl-a with multispectral data from Rededge Micasense on board the UAV platform for the 2018 flight. The models were done with in situ Chl-a and the median of the reflectance values included in a buffer of 1.5 m. around the sampling point.

Index | Interc. (a) | Slope (b) | R^{2} | Pearson r | p Value | Sig. Code ^{1} |
---|---|---|---|---|---|---|

SABI | 90.672 | 29.021 | 0.9099 | 0.9539 | 0.0461 | ** |

KIVU | 102.730 | −81.120 | 0.9056 | 0.9516 | 0.0483 | ** |

NDCI | 97.096 | 31.118 | 0.308 | 0.5550 | 0.4450 | |

NDVI | 91.361 | 20.354 | 0.750 | 0.8660 | 0.1340 | |

2BDA | 81.740 | 9.060 | 0.8651 | 0.9301 | 0.0698 | * |

3BDA | 93.810 | 2.270 | 0.0079 | 0.0891 | 0.9109 | |

3BDA_MOD | 96.936 | 0.1305 | 0.4257 | 0.6524 | 0.3475 | |

2BDA_2 | 78.740 | 18.430 | 0.2772 | 0.5265 | 0.4735 | |

B3B1 | 66.56 | 102.88 | 0.1709 | 0.4133 | 0.5867 | |

GB1 | 45.950 | 27.790 | 0.1644 | 0.4054 | 0.5945 |

^{1}(*) p < 0.1; (**) p < 0.05; (***) p < 0.01; (****) p < 0.001.

**Table 15.**Linear model coefficients and linear fit parameters for all the Spectral Band Combinations tested for the retrieval of chl-a with multispectral data from Rededge Micasense on board the UAV platform combining the data of 2017 and 2018 flights. The models were done with in situ chl-a and the median of the reflectance values included in a buffer of 0.5 m. around the sampling point.

Index | Interc. (a) | Slope (b) | R^{2} | Pearson r | p Value | Sig. Code ^{1} |
---|---|---|---|---|---|---|

SABI | 36.210 | 1110.010 | 0.0658 | 0.2565 | 0.552 | |

NDCI | 1.447 | −431.077 | 0.2879 | −0.5365 | 0.1364 | |

NDVI | 38.470 | 71.950 | 0.0456 | 0.2137 | 0.5809 | |

2BDA | −3.8180 | 39.805 | 0.0780 | 0.2793 | 0.4666 | |

3BDA | −10.15 | −222.900 | 0.4587 | −0.6773 | 0.0450 | ** |

3BDA_MOD | 14.629 | −1.686 | 0.4288 | −0.6548 | 0.0550 | * |

2BDA_2 | 263.500 | −266.00 | 0.2773 | −0.5265 | 0.153 | |

B3B1 | −41.980 | 520.310 | 0.9816 | 0.9907 | 0.0000 | ***^{*} |

GB1 | −205.40 | 175.30 | 0.9779 | 0.9889 | 0.0000 | **** |

GR | −168.90 | 120.00 | 0.8654 | 0.9302 | 0.0002 | **** |

^{1}(*) p < 0.1; (**) p < 0.05; (***) p < 0.01; (****) p < 0.001.

**Table 16.**Selected algorithms applied to the images of the reservoir obtained with Landsat 8 OLI, Sentinel 2 MSI and Rededge Micasense on board UAV.

Satellite - Sensor | Algorithm |
---|---|

Landsat 8 - OLI | Ln Chl-a = (1.448 * NDVI) + 1.923 |

Sentinel 2 - MSI | Chl-a = (26.795 * 3BDA_2) + 7.134 |

UAV-Rededge- Classification | Chl-a = (562.71 * B3B1) − 47.849 |

UAV-Rededge- Low chl-a | Chl-a = (0.2615 * 3BDA_MOD) + 3.715 |

UAV-Rededge- High chl-a | Chl-a = (28.895 * SABI) + 90.672 |

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Cillero Castro, C.; Domínguez Gómez, J.A.; Delgado Martín, J.; Hinojo Sánchez, B.A.; Cereijo Arango, J.L.; Cheda Tuya, F.A.; Díaz-Varela, R.
An UAV and Satellite Multispectral Data Approach to Monitor Water Quality in Small Reservoirs. *Remote Sens.* **2020**, *12*, 1514.
https://doi.org/10.3390/rs12091514

**AMA Style**

Cillero Castro C, Domínguez Gómez JA, Delgado Martín J, Hinojo Sánchez BA, Cereijo Arango JL, Cheda Tuya FA, Díaz-Varela R.
An UAV and Satellite Multispectral Data Approach to Monitor Water Quality in Small Reservoirs. *Remote Sensing*. 2020; 12(9):1514.
https://doi.org/10.3390/rs12091514

**Chicago/Turabian Style**

Cillero Castro, Carmen, Jose Antonio Domínguez Gómez, Jordi Delgado Martín, Boris Alejandro Hinojo Sánchez, Jose Luis Cereijo Arango, Federico Andrés Cheda Tuya, and Ramon Díaz-Varela.
2020. "An UAV and Satellite Multispectral Data Approach to Monitor Water Quality in Small Reservoirs" *Remote Sensing* 12, no. 9: 1514.
https://doi.org/10.3390/rs12091514