Spatial sampling for mapping

CropperCapture[323].jpg

Sound sampling design is essential for the collection of data to support reliable scientific inference and decision making for management and policy.  What counts as a sound design depends on the problem of interest, and the nature of the inference that is required.  Many statistical analyses of spatial data, particularly for the prediction of local values as in mapping, are done in a model-based context in which data are treated as realizations of an underlying random process.  In this setting it is not necessary to select sampling locations by probability sampling, and there is scope to optimize sampling patterns computationally.
In this course I shall introduce some of the concepts that underlay the optimization of spatial sampling.  These include methods to ensure good spatial coverage by a sample, experimental designs adapted for spatial surveys, model-based optimization of the grid-spacing and model-based optimization of the coordinates of sampling locations for ordinary kriging and kriging with an external drift. 

TIMETABLE

  • 9.00 - 9.30 Introduction sampling for mapping
  • 9.30 - 10.15 Spatial coverage sampling
  • 10.15 - 10.45 Break
  • 10.45 - 11.30  Fuzzy k-means sampling
  • 11.30 - 12.30  Response surface sampling and latin hypercube sampling
  • 12.30 - 13.30 Lunch
  • 13.30 - 14.30  Model-based optimization of grid-spacing
  • 14.30 - 15.00 Break
  • 15.00 - 16.00 Model-based optimization of coordinates of sampling locations
  • 16.00 - 17.00 Sampling for validation

requirements

Participants in the course will be provided with scripts for the free R platform which will allow them to use the  methods that are described to solve sampling optimization problems. Basic knowledge of R is required. 

Participants are encouraged to bring their own laptops with pre-installed R and if preferred, a GUI for R such as R Studio or Tinn-R.


Lecturer

Brus.jpg

Dr Dick J. Brus is senior scientist at Wageningen University and Research in the Netherlands and adjunct Professor at the School of Geography of Nanjing Normal University. He has rich experience in geostatistics and statistical sampling in space and time. He applies these statistical methodsin mapping and monitoring of natural resources such as soil, water and vegetation. He published numerous papers in international journals and is second co-author of the book Sampling for Natural Resource Monitoring (J.J. de Gruijter, D.J. Brus et al., 2006, Springer Verlag). In 2014 he was awarded the Visiting Professorship for Senior International Scientists by the Chinese Academy of Science.