Modelling Bluetongue vectors occurence using GIS and Remote Sensing techniques

828 views

Published on

Published in: Technology
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
828
On SlideShare
0
From Embeds
0
Number of Embeds
18
Actions
Shares
0
Downloads
0
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Modelling Bluetongue vectors occurence using GIS and Remote Sensing techniques

  1. 1. Modelling Bluetongue vectors occurence using GIS and Remote Sensing techniques Martins H., Nunes T., Boinas F.
  2. 2. Introduction <ul><li>Objectives </li></ul><ul><ul><li>Spatial modelling of Culicoides imicola occurence </li></ul></ul><ul><ul><li>Compare modelling techniques (discriminant analysis vs logistic regression) </li></ul></ul>
  3. 3. Materials and Methods <ul><li>Base data for modelling </li></ul>Temperature Eggs Larvae Animal density Temperature Humidity <ul><li>Entomological data </li></ul><ul><li>Climatic and </li></ul><ul><li>- environmental data </li></ul><ul><li>Animal population data </li></ul>
  4. 4. Materials and Methods <ul><li>Entomological data </li></ul><ul><li>- Ouput variable </li></ul><ul><li>- Entomological surveillance program </li></ul><ul><ul><ul><li>May 2005 </li></ul></ul></ul><ul><ul><ul><li>3670 valid catches </li></ul></ul></ul><ul><ul><ul><li>216 sampling locations </li></ul></ul></ul><ul><li>- Data processing criteria </li></ul><ul><ul><ul><li>Maximum count of individual </li></ul></ul></ul><ul><ul><ul><li>2 valid catches (September – November) </li></ul></ul></ul><ul><ul><ul><li>Treshold value = 10 specimens </li></ul></ul></ul><ul><ul><ul><li>Boolean classification (73 neg, 47 pos) </li></ul></ul></ul>Absence Presence
  5. 5. Material and Methods <ul><li>Climatic and Environmental data </li></ul>BIO1 = Annual Mean Temperature BIO2 = Mean Diurnal Range (Mean of monthly (max temp - min temp)) BIO3 = Isothermality (P2/P7) (* 100) BIO4 = Temperature Seasonality (standard deviation *100) BIO5 = Max Temperature of Warmest Month BIO6 = Min Temperature of Coldest Month BIO7 = Temperature Annual Range (P5-P6) BIO8 = Mean Temperature of Wettest Quarter BIO9 = Mean Temperature of Driest Quarter BIO10 = Mean Temperature of Warmest Quarter BIO11 = Mean Temperature of Coldest Quarter BIO12 = Annual Precipitation BIO13 = Precipitation of Wettest Month BIO14 = Precipitation of Driest Month BIO15 = Precipitation Seasonality (Coefficient of Variation) BIO16 = Precipitation of Wettest Quarter BIO17 = Precipitation of Driest Quarter BIO18 = Precipitation of Warmest Quarter BIO19 = Precipitation of Coldest Quarter WorldClim package
  6. 6. Material and Methods <ul><li>Climatic and Environmental data </li></ul><ul><li>- 30 remotely sensed variables </li></ul><ul><li>- NOAA satellite, AVHRR sensor </li></ul><ul><ul><ul><li>NDVI </li></ul></ul></ul><ul><ul><ul><li>LST </li></ul></ul></ul><ul><ul><ul><li>MIR </li></ul></ul></ul>Temporal encapsulation through Fourier trasnformation Maximum Minimum Average Amplitude Phase Anual, bi-anual, tri-anual cycles
  7. 7. Material and Methods <ul><li>Climatic and Environmental data </li></ul><ul><li>- SRTM sensor </li></ul>Elevation Slope Slope extension Aspect
  8. 8. Material and Methods <ul><li>Animal population data </li></ul><ul><li>- Bovines (SNIRB, DGV) </li></ul><ul><li>- Small ruminants (SNIRA, DGV) </li></ul><ul><li>- Kernel density </li></ul>Bovines Small Ruminants Ruminants
  9. 9. Material and Methods <ul><li>Conceptual Framework </li></ul>
  10. 10. Results <ul><li>Discriminant analysis </li></ul><ul><li>- Significant variables </li></ul><ul><ul><ul><li>Mean Temperature of warmest quarter (-) </li></ul></ul></ul><ul><ul><ul><li>Minimum LST (+) </li></ul></ul></ul><ul><ul><ul><li>Maximum MIR (+) </li></ul></ul></ul><ul><ul><ul><li>Tri-annual LST amplitude (+) </li></ul></ul></ul><ul><li>- Accuracy assessment </li></ul><ul><ul><ul><li>Se = 76,6% </li></ul></ul></ul><ul><ul><ul><li>Sp = 75,3% </li></ul></ul></ul><ul><ul><ul><li>Global Accuracy = 75,8% </li></ul></ul></ul>Absence Presence
  11. 11. Results <ul><li>Logistic Regression </li></ul><ul><li>- Significant variables </li></ul><ul><ul><ul><li>Mean Temperature of warmest quarter (-) </li></ul></ul></ul><ul><ul><ul><li>Precipitation of wettest quarter (-) </li></ul></ul></ul><ul><ul><ul><li>Minimum NDVI (+) </li></ul></ul></ul><ul><ul><ul><li>Slope (-) </li></ul></ul></ul><ul><ul><ul><li>Bi-annual MIR phase (-) </li></ul></ul></ul><ul><ul><ul><li>Tri-annual LST amplitude (+) </li></ul></ul></ul><ul><ul><ul><li>Minimum LST (+) </li></ul></ul></ul><ul><ul><ul><li>Annual MIR amplitude (+) </li></ul></ul></ul><ul><ul><ul><li>Mean temperature of driest quarter (+) </li></ul></ul></ul><ul><li>- Accuracy assessment </li></ul><ul><ul><ul><li>Se = 80,9% </li></ul></ul></ul><ul><ul><ul><li>Sp = 83,6% </li></ul></ul></ul><ul><ul><ul><li>Global Accuracy = 82,5% </li></ul></ul></ul>Probability
  12. 12. Results Absence Presence Probability BTV1 and 4 outbreaks (2006, 2007) Methodology Class Temp. Warmest quarter Min. LST Slope Min. NDVI Discriminant analysis Absent 20,1 2778,4 7,3 1246,6 Present 23,5 2837,8 3,4 1345,7 Logistic regression Absent 20,5 1946,9 4,8 919,1 Present 23,6 2839,0 2,9 1264,8
  13. 13. Conclusions <ul><li>Logistic regression models were more robust and accurate </li></ul><ul><li>Temperature variables were significant in both models </li></ul><ul><li>Future developments: </li></ul><ul><ul><ul><li>Include other remote sensing variables </li></ul></ul></ul><ul><ul><ul><li>Use higher resolution images (spatial, temporal and spectral) </li></ul></ul></ul><ul><ul><ul><li>Repeat methodologies with time-dependent dataset </li></ul></ul></ul><ul><ul><ul><li>Other methodologies (variable selection, modelling) </li></ul></ul></ul>
  14. 14. Work in progress

×