CALGARY, Alberta -
Feb. 1, 2018 -
PRLog -- Maapera Analytics Inc. is excited to have participated in supporting the CLRA lunch and learn series on January 31st, 2018.
Big data and machine learning are transforming a number of industries, and the environmental sector is no exception. Improvements in data processing and sensor technology have enabled monitoring and high resolution data acquisition options previously thought to be unachievable. Specifically, machine learning and new sensor technology can be used to support land stewardship, remediation and reclamation activities.
Obtaining analytical data during remediation activities represents a significant cost for most projects, and high resolution site characterization is not possible without innovative quantitative field screening technology. A technological solution to this problem is the use of short wave infrared (SWIR) reflectance spectroscopy combined with machine learning to identify distinct spectral signatures for petroleum hydrocarbons (PHCs) in soil. As a result, a detailed three dimensional model can be generated for a site rapidly, supporting field decision making and reducing the risk of residual contamination left on site.
This technology has been deployed on a number of sites across Canada by Maapera to vastly increase data acquisition for environmental assessment and remediation projects, and results were compared to standard analytical laboratory methods. Overall, the results showed that the technology had false positive and false negative rates less than 5 percent. Additionally, the overall goodness-of-
fit between field obtained data and lab data was over 90%. These results show that quantitative, reliable and rapid field screening is possible. It is now possible to go from a handful of data points for a site, to a cost-effective three dimensional model with a vertical resolution as low as 10 cm.
Additionally, machine learning and big data can support reclamation and land management activities. Soil carbon and nitrogen measurements have been successfully made using in-field SWIR reflectance spectroscopy methods. Due to the relatively lower cost compared to conventional laboratory methods, data can be obtained at much higher volumes. As a result, quantitative soil carbon monitoring is possible for a number of applications such as carbon offset verification or reclamation applications.
The development of machine learning and big data has to potential to transform environmental management. High volumes of quantitative data can now be generated and analyzed, reducing uncertainty for environmental management problems, and supporting managers to make better land stewardship decisions.
https://www.clra.ca/
Account Phone Number