Mapping Land Cover and Insect Outbreaks at Test Sites in East Siberia

Mapping land cover and insect outbreaks at test sites in East Siberia

V.I. Kharuk V. N. Sukachev Institute of Forest, Krasnoyarsk, Russia Insect outbreaks are one of the primary factors of Siberian forests dynamics, determining species composition and carbon balance. This table lists the main insect species which caused outbreaks in Siberian taiga Insect species Main food source tree species Max. outbreak area, thousand hectares

1. Dendrolimus superans sibiricus Tschetw.* fir, pine, larch, spruce > 1000 2. Lymantria dispar* larch, broadleaf 300 3. Orgyia antiqua larch, birch 40 4. Dasychira abietis spruce, fir, pine, larch 1000 5. Leucoma salicis* aspen, willow 100 6. Lymantria monacha* pine, spruce 5 7. Ectropis bistortata* spruce 400 8. Bupalus piniarius* pine 50 9. Semiothisa signaria* spruce, fir 10 10. Simiothisa continuaria larch 5 11. Erranis jacobsoni* larch 50 12. Biston betularius birch 50 13. Phalera bucephala birch 50 14. Clostera anastomosis* aspen, willow 10 15. Zeiraphera rufimitrana fir 100 16. Coleophera dahurica larch 100 17. Zeiraphera griseana larch >1000 The most dangerous species is the Siberian silkmoth (Dendrolimus superans sibiricus Tschetw.). This insect causes forest damage and mortality up to >1 mln ha per outbreak The map of actual and potential Siberian silkmoth food base

Outbreak areas: 1 (yellow line) – of 1950s; 2 (red line) - of 1990s, 3 (black circle) - of 21st century;

The temporal trend in outbreak periodicity is observing: 25 yr, 25yr, 12 yr, and 8 yr.

Solid line: northern limit of outbreak zone. Under the impact of climate change this border will shift northward. Secondary pests (bark beetles) attacked stands weakened by Siberian silkmoth Insect-killed stands becomes a “firewood” for the periodic follow-on fires The catastrophic scale of outbreaks requires adequate methods of detection and mapping.

The purpose of this study is evaluation of Terra/MODIS and SPOTVegetation data for detection of Siberian silkmoth outbreaks. zone of Siberian silkmoth outbreak in 1990’s. The disturbances: pest outbreak, logging, fires (including outbreak-promoting) A landscape: plateau InsectInsect outbreaksoutbreaks inin CentralCentral SiberiaSiberia

1950’s 1990’s

Insect outbreak

areas 2000’s

Landsat 7 image

AVHRR derived and conventional mapping insect damage Area of investigations _2: zone of Siberian silkmoth outbreak in 2000’s. The disturbances: pest outbreak, logging, fires SPOT Veg Image A landscape: mountains

1. The outbreak zone. 2. The main outbreak spot. NDVI difference dynamics between healthy and damaged stands (normalized data)

∆NDVI120

1 2 90

0 08/21/99 07/27/00 08/20/00 08/21/00 05/01/02 05/11/02 05/17/02 05/21/02 05/25/02 06/01/02 06/02/02 06/10/02 06/18/02 06/21/02 07/01/02 07/04/02 07/20/02 07/21/02 07/28/02 08/01/02 08/05/02 08/11/02 08/21/02 08/29/02 09/01/02 09/06/02 09/11/02 09/14/02 09/21/02 09/30/02 10/01/02 10/08/02 10/11/02 10/16/02 10/21/02 10/24/02 02/21/03 07/04/03 08/21/03 08/20/04 08/21/04 -30

1 – SPOTVeg 2 – TERRA/MODIS Damaged stands area dynamics

S, ha 20000 1 2 16000

12000

8000

4000

0 07/21/02 07/28/02 08/01/02 08/05/02 08/10/02 08/11/02 08/21/02 09/01/02 09/02/02 09/06/02 09/11/02 09/21/02 09/30/02 10/01/02 10/04/02 10/08/02 10/11/02 10/16/02 10/21/02 10/24/02 02/21/03 03/06/03 03/13/03 07/04/03 08/21/03 08/29/03 Date 1 – SPOT-Vegetation; 2 – Terra\MODIS Is the initial outbreak phase related to landscape features (aspect, slope steepness, and elevation)? To answer this question, we used high resolution 3D model N 25

NW NE 11/08/02 01/09/02

Damaged stands area increment 21/08/03 dynamics (with respect to aspect) W 0 E

SW SE

S 0.0042 S n

08/11/02-08/21/02

08/21/02-09/01/02 0.0021 Damaged stands area increment dynamics 09/01/02-09/11/02 (with respect to slope steepness)

0 1 6 11 16 21 α, degrees 26 ∆S2n S4n

08/11/02- 1 08/21/02 2 08/21/02- 09/01/02 09/01/02- 1 09/11/02 Healthy

0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

8 2 6 0 4 8 2 6 0 4 8 2 6 0 4 8

2 3 3 4 4 4 5 5 6 6 6 7 7 8 8 8

------

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

6 0 4 8 2 6 0 4 8 2 6 0 4 8 2 6

2 3 3 3 4 4 5 5 5 6 6 7 7 7 8 8 Altitude, m above sea level Increment dynamics of normalized damaged stands ( ∆Sn) with respect to altitude (Sn – normalized area of healthy stands) 3d view (not a map!) on areas damaged pests Conclusions

1. SPOT-Veg and Terra/MODIS sensors showed a good accuracy (>90%) for the outbreak area estimation.

2. Initial outbreak phase is related to topography features. This may be used for early detection of pest outbreak.

3. Mapping of pre- and post- outbreak temporal dynamics is necessary. ACKNOWLEDGMENTS The activity was supported by the NASA Science Mission Directorate and RFFI (Russian Fund for Fundamental Investigations) grant 03-05-65333 СПАСИБО ! THANK YOU!