Investigators
- Guido Cervone, Penn State University (Primary Investigator)
Description
Citizen-led movements producing scientific environmental information are increasingly common during hazards. After the Japanese earthquake-triggered tsunami in 2011, the governments of the United States and Japan collaborated to produce airborne remote sensing data of the radiation levels after the Fukushima nuclear crisis. Advances in technology enabled citizens to monitor on-ground levels of radiation by innovative mobile devices built from off-the-shelf components.
The citizen-led SAFECAST project recorded 50 million radiation measurements since 2011. The analysis of these data presents big data challenges because they are multi- dimensional, not vetted, and provided from multiple devices without quality control. This citizen science collection produces open radiation measurements at specific coordinates and times, yet the reliability and validity of the overall data had not been assessed. The nuclear disaster at the Fukushima Daiichi nuclear-power plant provides a case for assessing the SAFECAST data with official government data. A comparison of the dataset requires several preprocessing steps. The ionized radiation sensors were calibrated scientifically but collected using different units of measure. Radiation above background level decays over time in proportion to the decay rates of the elements present so integration of radiation surveys were necessary for temporal interpolation to compare measurements as being the same time frame. Finally, the GPS located points were selected within the overlapping extents at multiple spatial resolutions.
This study spatially analyzes and statistically compares citizen-volunteered and government-generated radiation data. Quantitative measures are used to assess the similarity and differences in the datasets. Radiation measurements from the same geographic extents show similar spatial variations which suggests that citizen science data can be comparable with government-generated measurements. Validation of SAFECAST would demonstrate that we can infer scientific data from unstructured and not vested data. Citizen science can provide real-time data for situational awareness which is crucial for decision making during disasters. This project provides a methodology for comparing datasets of radiological measurements over time and space. Integrating data for assessment from different earth sensing systems is an essential step to address big data challenges of volume, velocity, variety, and veracity.