Hyperspectral information for adoption and technology tracking

Transcript

1. Hyperspectral information for adoption and technology tracking Andy Nelson & Parvesh Chandna IRRI
2. How can spatial information augment M&E activities? IRRI obtained a small grant, 100k USD from SIAC ‘Strengthening Impact Assessment in CGIAR’ ! Advancing Methodologies for Tracking the Uptake and Adoption of Natural Resource Management Technologies in Agriculture Issued by: Michigan State University, on behalf of the SIAC Project of the CGIAR’s Standing Panel on Impact Assessment. July 29, 2013 ! • We are assessing hyperspectral information for discriminating between different management practices. • 12 month project, March 2014 – March 2015. Using field spectroradiometer in kharif and rabi season on rice and wheat in Haryana
3. How can spatial information augment M&E activities? • Before obtaining any hyperspectral imagery, we need to know what can be seen and in which part of the spectrum it can be seen. • Field work with a spectroradiometer to assess the spectral response of crops under different treatments.
4. How can spatial information augment M&E activities? • PSR-3500 High Resolution Full Range Portable Spectroradiometer • 350-2500nm spectral range • 512 element Si array and two 256 element extended InGaAs arrays for best resolution in NIR ranges • Spectral resolution 3nm (700nm); 8nm @ 1500nm; & 6nm @ 2100nm • Around $70,000 USD Don’t drop it in the paddy field!
6. How can spatial information augment M&E activities? Karnal, Haryana, Aug 2nd 2014
7. 40 days after transplantin
8. 80 days after transplantin
9. How can spatial information augment M&E activities? • Observations taken weekly though the season on rice crops undergoing different treatments that compare traditional and innovative practices. • Detection of spectral band(s) and time of season that uniquely identify the practice. • How to scale up this information to map the spatial extent and diffusion of practices?
10. Can we identify the same band combinations in the hyperspectral imagery and use that to map the spatial extent of the practice? Can we even get HS data? Repeated each season to track diffusion
11. Next steps • Complete rice and wheat crop field assessment until Mar 2015 looking at resource conservation technologies. • Purchase imagery and assess feasibility. • Expand to varietal detection under BMGF project STRASA III • Investigate pest and disease signatures • Contribute to a spectral/temporal library.