Detecting, understanding and reacting to extreme environmental events: integrating the potential of societal data, citizen science, Earth observation, and novel data analytic approaches


Session lead: Dr. Miguel Mahecha (MPI-BGC); Dr. Ursula Geßner (DLR); Dr. Ilona Otto (PIK)

Extreme climatic events are expected to intensify globally and affect the functioning of ecosystems. The expectable impacts on societal dynamics call for progress in detecting, monitoring and understanding extreme events and impacts. One may assume that the civil society, research communities and policy makers are well equipped with multiple data sources. For instance citizen observation, punctual information, administrative data and regional to global trans-boundary datasets based on Earth observations are often freely available. But in fact available data and analytic approaches are insufficiently exploited in this context and rather reflect disciplinary perspectives.

Aims and Scope

We explore if pressing problems and needs related to extreme event impacts could benefit from a joint mining of multiple source data streams using analytic methods across scientific and stakeholder communities. We aim at exploring where an overcoming of scale and/or thematic mismatches might open new perspectives for regional stakeholders, global policy makers, and science to better understand extreme event impacts on the socioecological systems. We bring together a stakeholder community concerned about extreme climate events and their societal and ecological impacts, which would benefit from a joint exploration of:

* Socio-economic data: Data on demographic, education, economic, health, migration, or human populations collected by national statistics bureaus, international organizations, and other public and private entities. These data form the backbone to analyze and understand the human impacts of extreme event, for instance societal impacts of extreme environmental events of the same physical magnitude will vary strongly according to density of human settlements and the quality of housing and infrastructure. We are, however, constrained by a lack of availability of long time-series and most of socio-economic data is available only at the national level resolution, what causes a temporal and spatial mismatch with bio-physical data. There are recent promising attempts to integrate data from various sources (e.g. socio-economic with remote sensing data) as well as to integrate data collected by various organizations and at various time steps to overcome these difficulties.

* Citizen science data: Citizen science has the great potential to deliver timely data on large spatio-temporal scales as well as fine resolution. In particular, when linked to issues of high societal relevance, such as catastrophes or environmental problems where adequate sensors or indicator schemes are missing, citizens data can become invaluable. Data streams can be mediated via social networks and communication platforms, and contributed from crowd sourcing data as well as from targeted experts from practice and management. Citizen Science approaches can help to frame questions, to collect data as well as contribute to interpreting remote sensing data, where automated processes cannot replace the capacities of human eye, link these to local knowledge and communicate via other channels than common scientific outlets.

* Remote sensing data: Satellite-based time series data have the potential to provide spatially and temporally consistent, transboundary information on different kind of extreme impacts on land ecosystems. Satellite Earth observation archives reach back to the 1970s, and are since then updated with continuously improved datasets of new and upcoming missions (e.g. Sentinel missions, Landsat- and MODIS-continuity missions). These data archives and data streams allow for example monitoring and assessing historic and current floods in urban and agricultural areas, and their impacts on housing, infrastructure and yields. Other applications in the context of climate extremes are the detection of agricultural droughts, fires, outbreaks of agricultural diseases, or the assessment of heat waves and their impact on human health.

* Environmental monitoring data (e.g. climate, ecological, and marine observations): Multiple in-situ observation networks monitor terrestrial and marine ecosystems as well as agricultural production rates. These observations allow disentangling the effects of climate variability from e.g. human innervation. One difficulty lies in the shortness of consistently available records, in their punctual/small-area information content, or in the fact that not all data are globally consistently reported to central repositories or national statistics.

Questions:
The workshop will systematically screen the matrix spanned by the two dimensions “application area” and “data collection, curation & analytic methods” to address the following questions:

  • How can multi-source data be integrated across thematic and scalar levels?
  • What are the most promising methodological approaches towards a synthesis of extreme events and their impacts?
  • What kind of requirements do exist for improving data sets (data acquisition schemes) which allow for a better integration in multi-source analyses of extreme events and their impacts?
  • Where are data and knowledge gaps that need to be filled?
  • Can an inter-disciplinary and multidimensional approach to extreme event detection be translated into an early warning system?


We will also discuss if new developments in trans-disciplinary, multi-source data analyses have the potential to support an improved understanding of the interactions of natural and social systems at different temporal and spatial scales, and if they can help to better understand vulnerabilities of socio-ecological systems to extreme climate events. Results of this brain-storming workshop will be summarized in a research statement that highlights the required innovation to putting the above questions (as well as additional topics that may come up) into practice.