Timely production of high quality and evidence based rapid risk/outbreak assessments in response to communicable disease threats is important not only in human, but also among animals.
Advances in computing power, together with the amount of data obtained from disease surveillance, registries, sensors or digital traces enable machine learning, complex system analytics and standard statistical tools as well as computer simulation to be applied to the field of Veterinary Epidemiology.
Our team member was visiting Australia (UNE) in middle of Novermber analyze animal contact networks (based on RFID tags) in poultry farms from an infectious disease modeling perspective i.e. to estimate critical detection times.
Andrzej has shown 3 case studies: 1) modelling, 2) infodemiology, 3) sensors:
Nowadays researchers widely use Internet media data to investigate the behavioural and affective dynamics of the public during COVID-19 pandemics. However, non-English European languages are highly underrepresented and other non-human health problems are not covered at all. Cross-sector communication, collaboration and knowledge exchange are still significant challenges for practical integration of human, animal, plants and environmental health issues. Adoption of One Health solutions can be divided into two pillars:
Infodemiology is very useful in understanding social perception during the pandemic by quantifying dynamics of interest (demand and supply of content) and discourse patterns. It plays a complmentary role to standard tools such as surveys and allows for the analysis in real time.
Infosurveillance could be useful for one public health decision makers in some specific areas such as predicting disease prevalence or early warning and source detection of health crisis event (also complmentary to standard tools such as disease surveillance and environmental monitoring system).
We hereby invite to attend a mini-workshop on media monitoring and analysis for One Health issues at the Institute for Veterinary Epidemiology and Biostatistics, FU Berlin organised by Prof. Vitaly Belik and Mr. Andrzej Jarynowski, as part of the DFG research (458528774) project „COVINT“.
Speakers and topics of workshop: Digital traces on the Internet from the One Health perspective
Vitaly Belik (Leader of Working Group System Modelling, Institute for Veterinary Epidemiology and Biostatistics FU Berlin)
Andrzej Jarynowski (computational epidemiologist, Institute for Veterinary Epidemiology and Biostatistics FU Berlin)
Daniel Płatek (postdoctoral fellow at Departments of Administrative, Economic and Social Sciences, Lund University/Assistant professor (incoming) at Institute of Political Studies of the Polish, Academy of Sciences in Warsaw)
Situation at Odra/Oder river shows a potential of media monitoring in infoveillance (early warning) as well as infodemiology (analysis of the discourse) within One Health context. Moreover it requires international cooperation. The problem appear between Opole Region Lower Silesia, then river passes large city of Wrocław, then move through Lubusz region it reaches the German (Branderburg)–Polish border, and then drains along the border through Western Pomeriana Province into Szczecin Lagoon shared between Poland Germany (Mecklenburg-Vorpommern) which is linked with the the Baltic Sea.
Using media monitoring and analysis tools, it is possible to trace the events of water state and mass fish die-offs on the Oder and its tributaries.
Timeliness and precision for early warning One health event detection and its importance (infoveillance) and its impact on the society (Infodemiology) from data published on the web is crucial for crisis management harm reduction. Our focus is the integration of human, animal, and ecosystem health with interactive computational social science and digital epidemiology approach.
Infectious diseases of animals ASF, COVID-19 humans, as well as food safety or the state of air mainly in Poland but also in some extent in Germany have previously marked their mark on the operation of sanitary, veterinary, plant protection and environmental inspections. In addition, together with the pharmaceutical and labor inspection (at least in part of their duties) and their military counterparts, they form a conglomerate of the so-called biomedical / One health inspections. Health, food, water, and environment are all wider topics with silos-specific and specialist concerns, the collaboration across sectors and disciplines (with social and computational science) will contribute to solve crises. In the event introduction of a new factor (natural or intended) CBRNE, the most important is the detection time. Until now (18.08), the factor or set of factors causing the ecological disaster on the Oder has not been established, however, in-depth analyzes were not started until 11.08, and in this article I indicate that it should take place on 05.08 at the latest.
Using content in Polish and German between 25.07-18.08.2022 with the help of monitors by Buzzsumo, Brand24, Twitter API, EventRegistry, Frazeo or Google trends, some results of One Health importance are discussed.
Media monitoring in Poland
To understand event, context must be introduced. The first information about dead fish on the Gliwice Canal (Oder tributary) appeared at the end of June (practically during the whole month of July discussion about individual dead fishes and the state of the Canal were happening). July was also a month of intensive discussion on the topic of deepening the Oder. The negative impact of regulation and, inter alia, the movement of contaminated sediment from the river bed came up (according to the media discussion).
Between the mouth of the Gliwice Canal and the weir in Lipki at the western end of the Opole Voivodeship and on the Nysa Kłodzka On the Oder, there were no mentions on social media or regional websites about dead fishes in July.
Moreover, round 25-27.07, residents, agricultural and ecological organizations of the riverside of Oder and its tributes complained about extremely low water levels in the rivers and the problem of water abstraction for irrigation of fields.
Let us focus on the digital traces in the area of the Lower Silesian Voivodeship in the initial phase, ie until about 08.08, which could help identify the factors causing the plague and limit the damage caused by the catastrophe.
On 27.07, the social media of Oława and the anglers‘ associations reported the first mentions of dead fishes on the border of the Opolskie and Lower Silesian Voivodeships. On the Oława news portals, full press articles appeared on 28.07. Also on 28.07, the Environmental Inspectorate in Wrocław set up a Twitter account and communicated on social media (Twitter, Facebook) about the results of the research. Due to the fact that so far the fish kill phenomenon was local and incidental, the Environmental Inspectorate in Wrocław did not take further steps apart from intensified monitoring.
29-31.07 media coverage in Oława media about fish kill was high due to the organised action of catching and disposing of dead fishes.
31.07 Lower Silesian regional media (including Radio Wrocław and Gazeta Wrocławska, TVP3 Wrocław) present their reports. Also on 31.07 a warning was issued on Silesian internet portals about a rapid increase in the state of the Oder River from Chałupki to Kędzierzyn Koźle (which would later appear as a theory of draining water from retention reservoirs).
Between 01-03.08 there was no mention of the appearance of the dead fishes outside angling forums (dead fishes appeared in the mainstream of the Oder in Wrocław, but not in quantities indicating an ecological disaster). The Wrocław Left-wing party members intervened in the public and political sphere.
On 04.08 there were reports from Głogów (about smell and single dead fishes).
It is worth noting that the first city through which the Oder River flows throughs (in the sense that the direct population has access to the shoreline and travels over bridges (without screens) after Wrocław is Głogów. Thus exposure to non-professional users of the Oder River was relatively low, so fish carcases could be overlooked. This lack of fish carcases occurrence and anomalites in water qiality in local media will be an important part of epizootiological investigation. It supported hypothesis about biological cause of the disaster (non-continuity of events in spatio-temporal picture).
Thus we can see that Lower Silesia as a first region of the disaster is great source of data about early states of the disaster .
With the help of semantic maps, we can see in what context people where searching for Oder. For instance the Oder („Odra“) is also a name of the disease: measles as well as name of multiple football clubs as „Odra Opole“. There are intensified queries (breakout in relative increase in search volume last weeks) about geography of the river („gdzie płynie Odra, przez jakie miasta przepływa“) and what is the reason of river poisoning („dlaczego/czym Odra jest zatruta“).
As the occurrence of death fishes moved forward with the current, the first reports from 06-08.08 in Lubusz were again coming from angling community. Fish kill search queries gain a lot of popularity in the Lubusz region (see the darker green on the map below). The lower part of Odra river there is much populated by fishes and other fauna, so in results the mass of death fishes were the most spectacular there.
During the weekend of 06-07.08, anglers discussed the large number of dead fishes on the so-called „dead Oder“ in various areas around Wrocław. Also during the weekend of 06-07.08, a discussion developed on the web portals and social media of Głogów.
We see also a strong interest in Oder river among the all regions affected directly by this ecological disaster. Please note, that Silesian region, through which Oder and its tributes flows, has low interest in the topic, as this region haven’t been affected by fish kills this time.
On 09.08 the mainstream media in Lubusz (Gazeta Lubuska, Radio Zachód etc.) publicise the issue (previously Lower Silesia media did it). On 09.08 nationwide media (Onet, Dziennik, TVPinfo, Interia, Wprost, SE) reported on the incident too.
Oder in Polish news started from a local event (Wrocław and Lubusz) to become national wide issue around 12-14.08, when prime minister and central government visited sites. Furthermore some topics as possible pollution with mercury (borrowed from German media) raised.
10.08 is further propagation on nationwide media and nationwide social media. Only now are nationwide environmental organisations getting involved. Thus, we cans see sharp increase in published material since 10.08. Political organisations from Western Poland have been activated.
It’s important to mention, that the most useful for describe the even term was fish kill till 11.08, afterwards Oder were more commonly used. This may be and effect that people become to threat this event as environmental disaster, which is more than only fish kill.
Another important aspect is social media coverage, because wave of interest in social media has different shape (its sharper and has maximum before traditional media), emotional load and topics coverages.
On 11.08 media storm erupts (mainly on social media) in Poland. We can see that the sharp peak of interest was on 12.08 on google trends and also a sharp peak on 13.08 on Polish Social Media. The peak in news was both 12.08 and16.08 (first day after long weekend – Bank holiday).
Keywords used by content authors are diverse with some several focuses on late reaction of local authorities, the restrictions introduces by voivodeship authorities on 12.08.
The information needs were also satisfied in Wikipedia, where we see similar peak on 12.08 as in Google trends and social media, because of SMS (short mobile message) alert in affected by disaster regions.
Further on, the media situation has already gained momentum and the Oder has become the number 1 topic on the Polish Internet and an important topic also in Germany (but never dominating).
Media monitoring Germany
On 10.08 fist mentions appeared on the German regional internet (as the wave of death fishers reached the German part of the Oder river).
There is a lot of critic towards Polish side, due to not informing Germans about the disaster. Climate change and weather were mentioned too (not so often, but link between disaster on Oder with climate change does not exist in Polish media at all). There is also a discussion on other fish kills event across Germany (in much smaller scale).
In German language Oder means or, so using single keyword Oder is not recommended, the only topic Oder will be consider for further analysis.
On 11.08 German regional broadcaster spread a rumour about contamination of water with Mercury). This rumour was officially cancelled by authorities on 13.08, but information become ‚alive‘ in media for a while.
Peak of interest in Germany in news was on 15.08 (however flat, while on social media on 14.08 (it was Sunday), so main traditional media had a day delay. Its seems that German discussion is less emotional than Polish (lower negative sentiment load).
We can see clearly peak on 11.08 evening on Twitter which was related to Mercury rumour from the news. Peak of interest on social media was on 13.08 in Poland, but in Germany on 14.08.
Geography of the interest in Germany is even more concentrated in a Oder regions than in Poland. Thus, there is much less interest in non-affected regions of Germany, than non-affected regions in Poland. It’s worth to mention, that in opposition to Poland, where the Oder river flows through multiple regions, it’s only a border river of a single state: Brandenburg .
We can see on Wikipedia queries how Fish kill event became Environmental disaster in naming.
It seems that interest in Poland is around 10-folds bigger than in Germany in Google queries. The peak of interest in Google Trends in Poland was observed on 12.08 (as it was a day that citizens of Odra region received communicates from authorities, but in Germany on 13.08.
Oder river in both German and Polish News has been discussed from reasons not related to investigated disaster too (as this river has important social and economic meaning for these countries).
Public asking Google needed mainly information about reasons of kill fish (warum Fischsterben in der Oder/ Welsche Fische sterben) and asking what do to (kann essen/ was tun).
The disaster has different names in Poland and in Germany. In Poland it’s mainly called “Katastrofa Ekologiczna” Ecological/Environmental Disaster, however in German internet its usually called “Fischsterben” (Fish kill), however the popularity of each terms was changing with time.
What should we take away from this lesson? The aim of this analysis is not primary to find the source of the disaster, but mostly find out what appeared in the media discourse and how this should be used. It is a signal report highlighting the role of real time media infosurveillance, which could have supplementary role to standard bio-chemical One Health surveillance.
One Health integrates the health of humans, domestic and wild animals, plants, and the wider environment (including ecosystems) are closely linked and inter-dependent, exactly as in Oder river disaster case study. We see how SMS alarm on 12.08 was a successful crisis communication tool in Poland. We have even seen that refugees and migrant community (counting up to 10% population of Western Poland) get interested, because we have observed also and increase in interest in Ukrainian language). It is worth mentioning, however, that in the media discourse there are hypotheses and theories (also propagated by scientific celebrities or people with professors‘ titles) completely unrealistic. For example, it was not possible to dilute the water by discharging water from retention or flood control reservoirs on the tributaries of the Oder, because due to the drought it was empty. Unfortunately, we noticed in the German and then repeatedly strengthened on the Polish side, pressure from journalists and politicians, who propagated by distorting unconfirmed information about mercury.
monitoring of local discussion forums on social media and local media as well as specialised discussion groups (in this case anglers) allows for spatio-temporal localisation of incidents in and should be incorporated into early warning/signalling systems and can help in the investigation of sources of possible contamination in Lower Silesia and Opole/Silesian Regions (infosurveillance);
discourse analysis on local and national social and traditional media allows for an understanding of threat perception and thematic dynamics (infodemic). We can see that authorities alerts works out (there was an important increase in interest). There is also a substantial difference in terms of intensity as well as topic narration between Poland and Germany (Fischsterben/Oder vs Odra/Katastrofa ekologiczna)