I would like to share my observations on the topic of environmental protection following natural disasters. Once in my professional career I analyzed the crisis management problem quite thoroughly, and I came to the conclusion that this topic was and still is too huge for the current approach to solving it. It seems to me that without AI methods it is not possible to investigate such problems, where we have to operate with a huge amount of data related to a multitude of visible and hidden dependencies.
There is a number of autonomously developed crisis management methods in Europe, e.g.
GMES (Global Monitoring for Environment and Security) the European Earth Observation program, launched in 1998 - monitoring of land, marine environment, atmosphere, climate change, security, and management in emergency situations; in particular, different catastrophic circumstances: floods, wildfires, landslides, earthquakes, volcanic eruptions and humanitarian crises.
GEM (Global Earthquake Model) - European project.
I can recall the NATO flood forecast project conducted in 2011 and 2012, as part of the Science for Peace and Security (SPS) program. This project was implemented at the Pripyat River, on the border between Ukraine and Belarus.
In Poland, there are several independently operating crisis management systems dealing with specific disasters. This causes the reaction to occur when the danger is already well underway. For example, the Emergency Management Center receives fire messages from residents or cameras. This information is often overdue, and the fire is already at full blast when the fire brigade arrives.
Project ISOK (pol. Informatyczny system osłony kraju przed nadzwyczajnymi zagrożeniami - ang. Informatic system of country protection against extraordinary threats", carried out in 2007-2013 mainly against flooding. However, in the summer of 2013, when there was a large flood in Poland, the ISOK system did not work properly, although it had full meteorological information. The water level forecast in ISOK was calculated based on various theories and water level monitoring, but the high rainfall in the mountainous regions of the country was not taken into account.
GMES, GEM and ISOK projects are powerful, IT-rich and able to solve the problems in their domain. Information from GMES is used by the respective projects, but GEM and ISOK are self-contained and do not share information. On the other hand, every great natural disaster is related to other disasters statistically, logically, or by causal relationship. For example, an earthquake can cause tsunamis, floods, fires, etc.
In crisis management, the task is to create a system that does not lose its functionality in hostile conditions. In other words, in the developed system, the security parameter is transformed into resilience according to the following scheme:
security -> dependability -> resilience
dependability - system performance parameter with the indicated trust when the critical infrastructure is destroyed (includes security as an attribute);
resilience - parameter that describes the system's ability to effectively accommodate unforeseen environmental perturbations or disturbances (includes dependability as an attribute).
Material loss (things, buildings, etc.) in the area of the disaster is not the only side effect. Natural disasters also cause changes in soil structures, often extending far beyond the area of the incident. This may lead to the collapse of the farming in this area. The probability of such a collapse is much greater when calculating only material losses (risk R is calculated as multiplying the probability P of an event for losses S as a result of this event: R = P x S) taking into account changes in soil structures because of the disaster. Analysis of soil changes and soil profile requires an assessment of its condition before and after the disaster. It is a great challenge nowadays.
So far, a huge amount of soil data has been collected, mainly on a local scale, for agricultural (and other) development and monitoring processes. At the same time, the consistency in the description, soil classification and analytical methods used were poor. In recent years, there is a growing consensus on each of these issues, and many national classifications have come forward in their approach and terminology.
The FAO Guidelines (FAO - Food And Agriculture Organization) for Describing the Soil Profile (FAO, 2006) and the USDA Soil Testing Manual (USDA - U.S. DEPARTMENT OF AGRICULTURE, 1993) are now recognized as international standards. In soil classification, the development of the World Reference Base of Soil Resources has resulted in many national classifications being converging in approaches and terminology. However, there is still no general agreement on a single system. There are significant advances in the standardization of soil laboratory methods, although texture and organic carbon measurements are still not standardized.
The visible reduction in central government funding in most countries of the world since the 1980s has shifted responsibilities from central soil and research organizations to regional groups and / or private sector organizations. Hence the lack of uniformity in the approach and methodology used, the proliferation of different soil classifications, the lack of availability of information after completion of the studies, and difficulties in harmonizing information at national and continental levels.
At regional level, SOTER (databases- Soil and Terrain) collected soil maps and soil profiles and organized them using a standardized methodology and soil classification system. This allowed some regional consistency of information for much of Africa, Europe, South America and the Caribbean. However, the lack of permanent funding calls into question the future of the SOTER program.
At a global level, the Harmonized World Soil Database brings together available information from various national and regional soil mapping programs such as DSMW (Digital Soil Map of the World), SOTER, China's National Soil Map and the European Geographic Database, and is currently the only global digital database of soil product available.
It should be emphasized, however, that most of the available global and regional data were obtained in the 1960s and 1990s. There is a significant gap between the production of a soil map in different regions today as there is no alternative soil information up-to-date currently available.
Previous soil surveys in most countries have collected up to millions of field observations and tens to hundreds of thousands of soil samples analyzed per country in the laboratory. Most of them have never been collected or published in open databases due to the considerable effort required to collect the data.
It seems to me that the time has come when the global soil science community must bring all its data from field observations and laboratory analyses to a centralized global center for the storage, managing, and sharing of soil data worldwide.
The regulation of soil data is the beginning of a variety of serious activities:
- increasingly sophisticated methods of processing huge amounts of data;
- extensive use of contextual information, contained by databases;
- data visualization;
- increasingly complex calculating methods of situation analysis and forecasting the situation in real time;
- the use of robots with remote control and AI elements for making independent decisions in specific situations and navigation in the indicated area;
- automatic connections of heterogeneous systems,
- extensive use of the theory of social behaviour for the analysis and correction of the situation;