Pre-Notes Notes

Schedule

please update in real-time throughout the day

Legend (feel free to modify)

• general notes look like this
• questions asked to presenter(s) look like this
• v important advice given to presenter(s) looks like this
• answers given by presenter(s) look like this
• questions asked by presenter(s) look like this
• action items look like this

Notes

Science

Notetaker: @Riaab Zahid

• General notes
  • requirements
    • Have a signal to noise ratio of at least 30:1
    • obtain images with sample distance ≤ 90m
    • capable of conducting hyperspectral imaging
  • objectives
    • take and downlink and image
    • downlink image of target
    • process downlink image
    • fit linear spectral modelling to image data
  • make decisions to fit all objectives na d model the design of the satelllite to helo complete phase c
  • working on building a model for crop residue
    • to see if requirements make sense - some major concerns
      • no satellites with SNR with 30:1
      • in literature spectral feature used is well outside the range of our sensor
  • may have problems mixing because soil and crop residue appea very similar
  • have made noise function
    • take spectral library entries and add noise function - take all library entries and re simulate data to see how FINCH will see it
  • Risk management
    • if SNR not viable can build algorithms into data processing pipeline and change the image capturing procedure
    • ground truth failure
    • selected components not optimal for scientific objectives
  • Next steps
    • rebuild science team
    • increase complexity of linear mixing model
    • choose abundance linear model
    • ground truth operations
• Questions to presenters
  • one of the risks may be that the SNR is too high - what is concern
    • SNR is too low for ours
  • How would SNR algorithms look like
    • the more data points we have of the same thing the more we do not need each pixel we can make data from multiple pixels
  • possible LMM not optimal can we figure this out in advance
    • no, not much literature review
  • are we using full range of the blue box
    • reduce amount of data which is not usable to help compression so we are not sure
  • do we have a general idea of how we will be making use of the spectral imaging
    • try to make a model - there is no literature
  • what determines the range for the requirenent
    • no documentation
  • what decides the spectral hardware requirements
    • higher level spacecraft departments decide - sensor range is blocking
  • what is the action is we cant use the sensor
    • rescope or find new sensor - has to occur in phase C
  • Where does science have to be to be comfortable to move to phase D
    • be bale to model design assembly and response to its objectives
  • What will you do when in phase D
    • improving the model
  • across band spectral imaging is diverging which shows crop residue
  • Do ew have efficiency graph of the sensor
    • we have 60 - 70 % in our range
• Add objective levels to existing mission objectives requirements

Systems

Notetaker: @Riaab Zahid