BloomTool

Abstract

The proliferation of algae (algal bloom) is an importantenvironmental issue as it affects a large number of living organisms, which andit is effects are directly or indirectly related with the aquatic environment.Algal bloom may result in suffering of fish and other aerobic organisms fromreduced oxygen and increased ammonia, which is extremely toxic. Depends on thetypes of algae the environmental impacts should increase. They are classifiedas harmful and can produce a wide range of toxins that can directly affectmarine animals and may also effect indirectly the consumers of them. Bloomsoccurs due to several factors, but the mainly due is the combination ofnutrient accumulation (nitrogen and phosphorus) in water, temperature and someextreme events. These factors measured and contrary to what is observed withregard to sensing by chemical and biological sampling methods. Satellite remotesensing provide data in a smaller time scale without additional cost andpractically unlimited range. Knowing the problems mentioned above, this projectproposes the use of remote sensing by spectrometry, by using new approachesthrough already operating equipment, to obtain nutrient concentration and watertemperature data that lead to the emergence of algal bloom. From these data,interpretations of them by artificial intelligence (AI) have the potential togenerate predictions about the occurrence or not of algal blooms in certainregions of the planet by sending alerts to any person/institution that haveinterest to be notified about the environment.

1. Justification and Objective

The exacerbated proliferation of algae (algal bloom) is an important environmental issue as it affects a large number of living organisms, which are directly or indirectly dependent on the aquatic environment. As it occurs, there is a change in the microorganism composition of the lower stratum in the water column, as other photosynthesizing organisms die due to lack of light, triggering the formation of an anoxic environment. At the end of this event, algae tend to die and settle to the bottom of the watercourse, providing more raw material for decomposing microorganisms, generating peaks in ammonia concentrations and worsening anoxia. As a result, fish and other aerobic organisms suffer from reduced oxygen and increased ammonia, which is extremely toxic.

Some types of algae aggravate environmental impacts even more. They are classified as harmful and can produce a wide range of toxins that can directly affect marine animals, bather and, indirectly, consumers of these animals. The main problems caused in humans are gastrointestinal (diarrhea) and neurological (paralysis, amnesia). Around 2,000 cases of food poisoning by algae toxins per year from ingestion of contaminated animals are reported worldwide, with a fatality rate of 15 % [3]. On the other hand, from an economic point of view, the losses are also significant. At times of flowering, riverside (freshwater), coastal (saltwater) and tourist populations are unable to access the affected watercourses, thereby stagnating all adjacent trade. In addition, in open environments, aquatic organisms affected by unfavorable conditions may flee elsewhere, however, for fish and other animals confined to fish farming for example, the consequences may be catastrophic.

Blooms occurs due to several factors, but mainly due to the combination of nutrient accumulation (nitrogen and phosphorus) in water, temperature and some extreme events. This factors can be measured and contrary to what is observed with regard to sensing by chemical and biological sampling methods, satellite remote sensing allows data to be obtained in a smaller time scale, without additional cost and practically unlimited range. Knowing the problems mentioned above, this project proposes the use of remote sensing by spectrometry, using new approaches through already operating equipment, to obtain nutrient concentration and water temperature data that lead to the emergence of algal bloom. From these data, interpretations of them by artificial intelligence (AI) have the potential to generate predictions about the occurrence or not of these blooms in certain regions of the planet, making it possible to carry out alerts to any person/institution intending to have contact with the environment to be affected.

2. Factors that causes harmful algal blooms

Algal bloom results from many factors such as wind, water currents, mean depth, mixing regime, water temperatures, and also extreme events (floods, hurricanes, and droughts and sluggish water circulation) [4]. Occurrence of blooms correlates positively with nutrients such as phosphorus, nitrogen, and carbon [9], and these correlative relationships have been used in TMDL (Total Maximum Daily Load) development. It has been proved that some specific concentration of this essential nutrients are needed to algal blooms, otherwise it doesn't happen [6,7]. Human activities have a huge impact in the increasing of these nutrients in water bodies and some examples include: runoff from agricultural, wastewater effluent (municipal and industrial) and runoff from mines [13].

Differences between the hydrology of water bodies may also influence bloom occurrence. For example, headwater streams are more limited by nitrogen than phosphorus, have fast flows, are narrow, shallow, and have majority of bentonic algae. By the other hand, riverine water have an increase in nutrient provided by human activities, they have slower flows, hence longer residence times, favoring algal blooms. In the ocean the main nitrogen loading occurs by precipitation, provided by atmospheric emission [10]. Also, ocean has lower temperature variation and high salinity.

3. Methodology

To solve the bloom problem it is necessary to create a prediction model. In this way, first step is to get data about the variables that precedes the bloom. There are several institutions that collect water samples, but these data are local and depend on available infrastructure. For this reason, a method of obtaining data by remote measurements may be more viable for larger scale forecasting, as well as avoiding intrusive tools, bypassing problems such as pollution in tool installation and maintenance processes.

The remote sensing process consists of obtaining and recording data without physical contact, using electromagnetic radiation through tools such as scanners, cameras, antennas that can be mounted on mobile platforms such as airplanes, spacecraft and satellites. The effort directed to data processing and analysis is one of the main advantages of the method used, since the time limiting of presential sampling does not interfere in the process, besides the advantage that each satellite can cover a large geographical area. The use of satellite imagery has been used in a range of applications, such as physics, biology, biologic oceanography, meteorology, geology, animal tracking, and environmental monitoring.

Among all the cited applications, the aim of this work is the quantitative analysis of phosphorus, nitrogen, temperature, weather and ocean currents. These factors determine the algal outcrop process. Than, measures can be calculated by analyzing satellite image data (Figure 1 - See the link bit.ly/35OXDXP), as an example, we have LandSat, Kronos and Meteosat. Liu 2015, Torbick 2013 and Wu 2010 [7, 14, 15]. Analysis by MODIS standard images have found that it was possible to measure these parameters with reasonable precision, in which was compared with measurements obtained from water samples taken from the same region. In the Table (See the link bit.ly/2J6lqsI), we have shown studies that propose methods for predicting algal bloom using MODIS images [11].

It is reveled from the collected data that is possible to organize them in temporal series [1, 2, 5, 8, 12]. This series are constituted by the qualification performed in a specific time period to present. Using this historic, it is possible to perform prediction of levels attributes to phosphorus, nitrogen, temperature, weather and ocean currents via deep learning. Thus, the occurrence of algal bloom is plausible.

This information will be obtained through images in the MODIS standard and will be used within the "BloomTool" tool whose main purpose is bloom predictions based in matematical models. Within this tool, the user can choose a geographic region and check aspects, such as: water temperature, sea currents and concentration of phosphorus and nitrogen in real time. This information is a set of factors that determinate the occurrence of a bloom as shown in figure 2 (bit.ly/2MVIIm5)Proposal of a tool that provide the users a model that predicts algal bloom based on interpretation of obtained data via satélite, using artificial intelligence.. Also, the BloomTools will informs bloom epicenters and will provide the affected areas .

4. References

[1] Ni-Bin Chang, Zhemin Xuan, and Y Jeffrey Yang. Exploring spatiotemporal patterns of phosphorus concentrations in a coastal bay with modis images and machine learning models. Remote sensing of environment, 134:100–110, 2013.

[2] Jayesh Ganpat Ghatkar, Rakesh Kumar Singh, and Palanisamy Shanmugam. Classification of algal bloom species from remote sensing data using an extreme gradient boosted decision tree model. International Journal of Remote Sensing , pages 1–27, 2019.

[3] Gustaaf M Hallegraeff. Ocean climate change, phytoplankton community responses, and harmful algal blooms: a formidable predictive challenge 1.Journal of phycology, 46(2):220–235, 2010

[4] H Kenneth Hudnell. Cyanobacterial harmful algal blooms: state of the science and research needs , volume 619. Springer Science & BusinessMedia, 2008.

[5] Sita Karki, Mohamed Sultan, Racha Elkadiri, and Tamer Elbayoumi. Mapping and forecasting onsets of harmful algal blooms using modis data over coastal waters surrounding charlotte county, florida. Remote Sensing , 10(10):1656, 2018.

6] Xiong Li and Hao Wang. A stoichiometrically derived algal growth model and its global analysis. Math. Biosci. Eng , 7(4):825–836, 2010.

[7] Jiaming Liu, Yanjun Zhang, Di Yuan, and Xingyuan Song. Empirical estimation of total nitrogen and total phosphorus concentration of urban water bodies in china using high resolution ikonos multispectral imagery. Water , 7(11):6551–6573, 2015.

[8] Nathan F Manning, Yu-Chen Wang, Colleen M Long, Isabella Bertani, Michael J Sayers, Karl R Bosse, Robert A Shuchman, and Donald Scavia. Extending the forecast model: Predicting western lake erie harmful algal blooms at multiple spatial scales. Journal of Great Lakes Research , 45(3):587–595, 2019.

9] Hans W Paerl. Nuisance phytoplankton blooms in coastal, estuarine, and inland waters 1. Limnology and Oceanography , 33(4part2):823–843, 1988.

[10] Hans W Paerl, Timothy G Otten, and Raphael Kudela. Mitigating the expansion of harmful algal blooms across the freshwater-to-marine continuum, 2018.

11] Thomas S Pagano and Rodney M Durham. Moderate resolution imaging spectroradiometer (modis). In Sensor Systems for the Early Earth Observing System Platforms , volume 1939, pages 2–17. International Society for Optics and Photonics, 1993.

[12] Benjamin P Page, Abhishek Kumar, and Deepak R Mishra. A novel cross-satellite based assessment of the spatio-temporal development of a cyanobacterial harmful algal bloom. International journal of applied earth observation and geoinformation , 66:69–81, 2018.

[13] M Romeo Singh and Asha Gupta. Water pollution-sources, effects and control, 2016.

[14] Nathan Torbick, Sarah Hession, Stephen Hagen, Narumon Wiangwang, Brian Becker, and Jiaguo Qi. Mapping inland lake water quality across the lower peninsula of michigan using landsat tm imagery. International journal of remote sensing , 34(21):7607–7624, 2013.

[15] Chunfa Wu, Jiaping Wu, Jiaguo Qi, Lisu Zhang, Huiqing Huang, Liping Lou, and Yingxu Chen. Empirical estimation of total phosphorus concentration in the mainstream of the qiantang river in china using landsat tm data. International Journal of Remote Sensing , 31(9):2309–2324, 2010.