Algorithms for HAB detection
A rapid increase in the number of microalgae is considered as an algal bloom. Harmful algal bloom (HAB) is a bloom that results in negative impact on plants and animals. Over the past decade, significant achievements were made to synoptically detect and characterize the location and extent of HABs from ocean color satellite sensors. A lot of greatly improved algorithms have been developed to more accurately retrieve phytoplankton proxies in coastal waters. Therefore, the SatCO2 software includes 5 HAB detection modules which have been successfully in some typical coastal regions such as the East China Sea, the Gulf of Mexico, and so on. The detailed descriptions of these modules are as following.

S7.1. The HAB algorithm in the East China Sea

This module integrates two HAB algorithms in the East China Sea. The first algorithm is a multispectral approach for discriminating P. donghaiense blooms from other water types based on MODIS Rrs spectral shape discrimination. Its procedure is separated into two steps. First, the bloom waters are identified by the lowRrs(555) and high RAB. Second, two new indices of P. donghaiense index (PDI) and diatom index (DI) are developed for discriminating P. donghaiense from diatom blooms. The second algorithm is a VIIRS based approach for detecting HAB waters. The detailed method is schematically illustrated in Fig S2.
(1)MODIS HAB algorithm description
Algal bloom ratio (RAB) is defined as:

P. donghaiense index(PDI):


Diatom index(DI):

(2)VIIRS HAB algorithm description
Algal bloom ratio (RAB) is defined as:

Bloom index (BI) is defined as:

Fig. S2 Schematic procedure of MODIS multispectral method for the identification of
P. donghaiense.

S7.2. The RBD based HAB algorithm

This module integrates the technique which was developed by Amin et al., 2009 primarily for detecting Karenia brevis (K. brevis) blooms throughout the Gulf of Mexico. This detection technique for blooms with low backscatter, which we name the Red Band Difference (RBD) technique These techniques take advantage of the relatively high solar induced chlorophyll fluorescence and low backscattering of K. brevis blooms. The techniques are applied to the detection and classification of K. brevis blooms from Moderate Resolution Imaging Spectroradiometer (MODIS) ocean color measurements.
(1)MODIS RBD algorithm description
Red Band Difference (RBD) is defined as:

Threshold of Algal bloom waters:
RBD > 0.15W/m2/μm/sr

Fig.S3 MODIS (Aqua) bloom image from 13 November 2004 for the WFS (a) FLH
(W/m2/μm/sr) image, (b) RBD (W/m2/μm/sr) image and (c) Normalized-water leaving radiance spectra taken from the bloomed and turbid waters indicated by “circle” and “squares” respectively in the FLH image.

S7.3. The RGCI based HAB algorithm

This module integrates a novel empirical Chla algorithm based on a Red-Green-Chorophyll-Index (RGCI) which was developed by Le et al., 2013, and validated for MODIS and VIIRS observations. The algorithm showed robust performance in coastal waters such as Tampa Bay and the northeastern Gulf of Mexico. Oi et al.(2015) used RGCI from VIIRS to observe Karenia brevis blooms in the Northeastern Gulf of Mexico, which can overcome deficiency rising from absence of a Fluorescence Band.
(1)MODIS RGCI algorithm description
Red-Green-Chorophyll-Index (RGCI) is defined as:
RGCI = Rrs(667)/ Rrs (547)
The empirical Chla algorithm:
Chla = 0.86 ∗ exp(5.1 ∗ RGCI)
(2)VIIRS RGCI algorithm description
Red-Green-Chorophyll-Index (RGCI) is defined as:
RGCI = Rrs(672)/ Rrs (551)
The empirical Chla algorithm:
Chla = 0.10 ∗ exp(11.8 ∗ RGCI)
Fig.S4 Comparisons between (a and d) VIIRS OC3 Chla (mg m−3), (b and e) VIIRS RGCI Chla (mg m−3), and (c and f) MODISA nFLH (mW cm−2 μm−1 sr−1) for (a–c) July 30, 2014, and (d–f) August 27, 2014. The annotate times are GMT hours and minutes. Black is land, and gray represents clouds or invalid data.

S7.4. The ABI based HAB algorithm

This module integrates new bio-optical algorithm has been developed by Shanmugam et al., 2011, to provide accurate assessments of chlorophyll a (Chl a) concentration for detection and mapping of algal blooms from satellite data in optically complex waters, where the presence of suspended sediments and dissolved substances can interfere with phytoplankton signal and thus confound conventional band ratio algorithms. This algorithm that uses the normalized water‐leaving radiance ratios along with an algal bloom index (ABI) between three visible bands to determine Chl a concentrations, which derived using Moderate Resolution Imaging Spectroradiometer (MODIS) data and can be could be extensively used for quantitative and operational monitoring of algal blooms in various regional and global waters.
(1)MODIS ABI algorithm description
Algal bloom index (ABI) is defined as:

The empirical Chla algorithm:

Fig.S5 (top) A MODIS/Aqua true color composite on 18 February 2010 in the Arabian Sea and Gulf of Oman. The corresponding Chl a images derived using (middle) the OC3 and (bottom) ABI algorithms. The combined SeaDAS algorithm and CCS scheme was used for atmospheric correction of MODIS/Aqua data.

S7.5. The FLH based HAB algorithm

This module integrates a MODIS fluorescence line height (FLH) based algorithm developed by Hu et al., 2005 to detect and trace a harmful algal bloom (HAB), or red tide, in SW Florida coastal waters. Because MODIS fluorescence line height (FLH in W/m2/μm/sr) data showed the highest correlation with near-concurrent in situ chlorophyll-a concentration (Chl in mg m-3, the results show that the MODIS FLH data provide an unprecedented tool for research and managers to study and monitor algal blooms in coastal environments.
(1)MODIS FLH based Chla algorithm description
The FLH based Chla algorithm:
Chl=1.255 * (FLH * 10)0.86

Fig.S6. MODIS/Aqua imagery for SW Florida coastal waters. Left column: Fluorescence Line Height (FLH; W/m2/μm/sr); the color scale includes negative values. Middle column: band-ratio chlorophyll concentration (OC3M Chl; mg m-3). Right column: enhanced RGB (ERGB) composite images from water-leaving radiance in three MODIS wavelengths: 551 nm (R), 488 nm (G), and 443 nm (B).
[1]     Tao Bangyi, Mao Zhihua, Lei Hui, et al. A semianalytical MERIS green-red band algorithm for identifying phytoplankton bloom types in the East China Sea[J]. Journal of Geophysical Research: Oceans, 2017,122(3):1772-1788.
[2]     Tao Bangyi, Mao Zhihua, Lei Hui, Pan Delu, ShenYuzhang, Bai Yan, Zhu Qiankun, Li Zhien, A novel method for discriminating Prorocentrum donghaiense from diatom blooms in the East China Sea using MODIS measurements[J]. Remote Sensing of Environment, 2015,158(0):267-280.
[3]     Tao Bangyi, Mao Zhihua, Pan Delu, ShenYuzhang, Zhu Qiankun, Chen Jianyu, Influence of bio-optical parameter variability on the reflectance peak position in the red band of algal bloom waters[J]. Ecological Informatics, 2013,16(0):17-24.
[4]     Tao Bangyi, Mao Zhihua, Pan Delu, ShenYuzhang, Optical detection of Prorocentrum donghaiense blooms based on multispectral reflectances[J]. Acta Oceanologica Sinica. 2013,32(10):48-56.
[5]     Ruhul Amin, Jing Zhou, Alex Gilerson, et al. Novel optical techniques for detecting and classifying toxic dinoflagellate Karenia brevis blooms using satellite imagery [J]. Optics Express, 2009,17(11): 9126.
[6]     Chengfeng Le, Chuanmin Hu, David English, et al. Climate-driven chlorophyll-a changes in a turbid estuary: Observations from satellites and implications for management [J]. Remote Sensing of Environment, 2013,130: 11-24.
[7]     Lin Qi, Chuanmin Hu, Jennifer Cannizzaro, et al. VIIRS Observations of a Karenia brevis Bloom in the Northeastern Gulf of Mexico in the Absence of a Fluorescence Band [J]. IEEE GEOSCIENCE AND REMOTE SENSING LETTERS, 2015,12(11): 2213-2217.
[8]     Palanisamy Shanmugam. A new bio-optical algorithm for the remote sensing of algal blooms in complex ocean waters [J]. JOURNAL OF GEOPHYSICAL RESEARCH, 2011, 116:C04016.
[9]     Chuanmin Hu, Frank E. Muller-Karger, Charles (Judd) Taylor et al. Red tide detection and tracing using MODIS fluorescence data: A regional example in SW Florida coastal waters [J]. Remote Sensing of Environment, 2005, 97:311-321.