MEYER (PDF)

Towards  an  Integrated  Early  
Warning  System  for  detec7ng  the  
onset  of  Deser7fica7on  in  Libya  
Pauline Meyer
Supervisor: Stephen Hallett
Context

 

Introduc7on:   Deser;fica;on   represents  
a  serious  environmental  problem  in  arid  
and   semi-­‐arid   zones,   caused   by   a  
Study  area  
c o m b i n a ; o n   o f   c l i m a ; c   a n d  
anthropogenic  factors.        
Deser;fica;on   cons;tutes   an   obstacle   to   urbanisa;on   and  
agricultural   development,   and   impacts   on   the   ac;vi;es   in  
Jeffara   plain,   Libya.   The   region   also   experiences   the   worst-­‐case  
for  water-­‐balance  of  the  country  due  to  Human  pressures.  
Aim:   Describe   a   GIS   model   approach   to   evaluate   the   sensi;vity  
to  deser;fica;on  and  to  analyse  its  spa;o-­‐temporal  evolu;on.  
Objec7ves:    
§  Iden;fy   and   integrate   the   driving   factors   of   deser;fica;on  
into  the  model  using  ArcGIS  
§  Analyse  spa;al  and  temporal  trend  using  Remote  Sensing  
technique  and  MODIS  NDVI  ;me  series    data  
§  Model  applicable  to  Middle  East  and  North  African  states    
§  Classify  the  results  into  deser;fica;on  risk  levels  

Methodology  
§  Model  approach  and  its  components  (MEDALUS  method):  
Rainfall  
Aridity  index  

Climate  
Quality  Index  
 
 
Vegeta7on    
Quality  Index  
 
 
 
Soil    
Quality  Index  

Fire  risk  
Erosion  protec;on  
Drought  resistance  
Plant  cover  
Parent  material  
Soil  texture  
Soil  depth  
Slope  gradient  
50,0  
40,0  

 

48,5  
38,8  

Area  (%)  

Results  

Sensi7vity  to  
Deser7fica7on  
Index  

Slightly  sensi;ve  
Sensi;ve  
Very  Sensi;ve  

30,0  
20,0  

§  Model  results:  

12,6  

10,0  

 

0,0  

SDI  =  (CQI*VQI*SQI)1/3  
(Sensi;vity  to  Deser;fica;on  Index)  

Slightly  sensi;ve  

Sensi;ve  

Very  Sensi;ve  

Methodology  

§  Assessment   of   MODIS   NDVI*   ;me   series   data   products   for   detec;ng  
deser;fica;on  spa;al  trend:    
MODIS  Terra  Vegeta7on  Indices    
acquisi;on  (250m  16-­‐day  NDVI)    
*Normalized  Difference  Vegeta;on  Index    

Results  

Random    
sampling  

Spa;al  trend  
analysis  

Pixel  reliability  
check  

Local  Polynomial  
Interpola7on  

§  Time  series  analysis:  

Discussion  and  Conclusions  

Further  work:    
 
§  Include   the   socio-­‐economic   index   due   to   significant   pressures   from   human   ac;vity   in   the  
The   vegeta;on   index   tends   to   be   the   most   determinant   factor   of   sensi;vity   to   deser;fica;on   in   this  
region  
study.   The   ;me   series   analysis   shows   a   spa;al   loss   of   vegeta;on   over   years   during   the   dry   season.   These  
§  Use  of  photogrammetry  for  vegeta;on  mapping    in  order  to  produce  a  more  precise  index  
results   validate   the   GIS   model   developed   in   Jeffara   plain.   Coastal   areas   tend   to   be   less   sensi;ve   to  
§  Use  of  new  satellite  plakorms  such  as  Copernicus  with  fast  revisit  ;mes  can  allow  improved  
deser;fica;on  than  the  steppes  in  the  Southern  part  where  the  loss  of  vegeta;on  is  clearly  detectable.      
method  
Key  literature  references:  Kosmas  et  al.  (1999)  “The  MEDALUS  project”  ;  Kibblewhite  et  al.  (2007)  “Environmental  Assessment  Of  Soil  For  Monitoring  Deser;fica;on  In  Europe”  ;  Salamani  et  al.  (2012)  “Evalua;on  de  
la  sensibilité  à  la  déser;fica;on  en  Algérie”  ;  Brandt  et  al.  (2014)  “Environmental  change  in  ;me  series  –  An  interdisciplinary  study  in  the  Sahel  of  Mali  and  Senegal”  

www.cranfield.ac.uk/courses/masters/geographical-informationmanagement.html
Cranfield University, Cranfield, Bedfordshire, MK43 0AL
Stephen Hallett - s.hallett@cranfield.ac.uk