Climate change is one of the most important challenges humanity faces at the moment. It is a phenomenon in which the climate of a particular region is modified and has a causality relation with both climate phenomena as well as with human behavior. Climate change has somewhat of a cyclic nature since it has a direct impact on humans and other living beings, leading to an increase in the amount of natural disasters, such as hurricanes, tornadoes, floods, and droughts. It can also lead to an increase of extreme weather conditions, such as heavy rainfall, melting glaciers, and rising sea levels. Understanding the impact climate change will have at the local level is therefore crucial for people to adopt steps to mitigate global warming and its foreseen consequences.
The NAIADES Water Observatory (access it through the NAIADES platform or at naiades.ijs.si) has already shown usefulness in understanding water-related topics as, e.g., extreme events, reclaimed water or water contamination, from the point of view of published science and worldwide news, and how could we derive best practices from its powerful text-mining capabilities. We can also use this innovative European-based technology to address climate change preparedness from its multiple angles, fed by news sources, scientific research innovation and the insight from patented technology. The evolution of the topic “Climate Change” across time (see Figure 1) can be obtained from the text mining of published scientific research and the main topics that were relating to it can be extracted to better understand the context of the scientific community achievements, helping us further to find new solutions for the forthcoming problems of this new era.
Figure 1 - The evolution of the topic “Climate Change” across time in the published scientific research and the main topics that were relating to it.
The new “Resources” view in the NAIADES Water Observatory aims to provide the user with a tool to predict the future river level and air temperatures. This has been done using mathematical models, based on the historical data. In the context of the workflow of the NAIADES Water Observatory we have analyzed over twenty years of daily environmental data at each of the three project use cases - Alicante (Spain), Braila (Romania) and Carouge (Switzerland) - to perform long-term ten year forecasts that allow us to identify trends that should be taken in consideration when planning for climate change preparedness. These predictions can be seen averaged by years, and further explored in the second visualization.
To analyze the Braila case, the NAIADES team had access to daily collected data since 2000 on average air temperature (degrees) and average air pressure (bars), sourced at Meteoromania, the Climatological Department of the Romanian National Meteorological Administration. The following charts show the forecasted trends over 10 years for the average air temperature and air pressure in Braila, Romania. We note that the difference between maximum and minimum temperature is reducing over time in our forecast, and that we are more prone to an increase in temperature values. To have a better understanding of what can happen in the future we used the the Koopman approach to understand the non-linear dynamics of the dynamical system where we can see accentuated max and min values (see figure 2).
Fig 2 - Koopman Braila Univariate Average Temperature
In Alicante we used sun hours and rainfall as features in the temperature and pressure forecasts. We obtained daily collected data since 2000, sourced at AEMET, the Spanish State Meteorological Agency, showcasing the min/max air temperature (degrees), the min/max air pressure (bars), the total sun hours (hours) and the precipitation (ml). The following charts show the forecasted trends over 10 years for the maximum air temperature and maximum air pressure (taking into account sun hours and precipitation) in Alicante, Spain (see Figure 3).
Fig 3 - Koopman Alicante Univariate Maximum Temperature with Features
In the case of Carouge we looked into the water temperature (degrees) and water levels (meters) daily collected data since 2002, sourced at the Meteo Swiss Data Portal IDAWEB. This provides us a different perspective on the impact of climate change. The following charts show the forecasted trends over 10 years for the average air temperature, water levels and water temperature in Carouge, Switzerland (see Figure 4).
Fig 4 - Koopman Carouge Univariate Water Temperature
A different perspective of this analysis of environmental parameters in the perspective of water, is done through Markov chains through the Streamstory technology. It allows us to examine different time-series simultaneously, and identify automatically generated states and the relation between them (from the input environmental data), that can provide us with a perspective on how seasons behave and are changing across the 20 years of acquired data at the climatological national agencies. When we just observe the Markov chain analysis of the water temperature and water levels, we already see the seasonality as distributed throughout the year. On the other hand, the analysis of temperature and pressure in Carouge shows different stages of low pressure to take into consideration, impacted by the scarce data provided by sun hours and precipitation (see Fig 5).