METHODS
I used ENVI Remote Sensing software to perform change detection on satellite imagery from three points in time:1. Ahmanson Ranch during bridge construction (mid-prep)
2. After the filming of Mission Impossible III (post-production) & Before the fire
3. After the fire
I downloaded free satellite imagery of Southern California from the Landsat 5 Thematic Mapper (TM) sensor at 30 meter spatial resolution using the NASA sponsored website Reverb/Echo. Construction of the bridge began in July 2005 and the filming took place at the end of August and beginning of September. A fire swept through the production site at the end of September. I downloaded one image from each month: August, September and October.
I then processed the three images (mid-prep, after production, after fire) with ENVI and cropped a spatial subset for the film location site – Ahmanson Ranch, Upper Las Virgenes Canyon Open Space Preserve, Ventura County (GPS: +34° 10' 30.03", -118° 40' 52.73") (images can be found in results section). I created a Normalized Difference Vegetation Index (NDVI) for each image. An NDVI measures the amount of natural vegetation and not only shows “greenness” but also calculates units of biomass in an area. Next I performed change detection to determine areas of changed vegetation. I compared how much the vegetation changed between mid-preparation and after the production and I compared how much the vegetation changed from after the production (before the fire) and after the fire. Red indicates an increase in vegetation and blue indicates a decrease in vegetation. A darker shade of either color corresponds to a larger increase or decrease of vegetation. I saved both change detections as .TIFF files and overlaid the images onto Google Earth Pro. The images can be viewed in the results section.RESULTS
I used ENVI Remote Sensing software to perform change detection on satellite imagery of Ahmanson Ranch from during site preparation and after the filming of Mission Impossible III. Red indicates an increase in vegetation and blue indicates a decrease in vegetation. A darker shade either red or blue corresponds to a more significant increase or decrease in vegetation. The series of change detection images on the left hand side show the difference in vegetation from during the bridge construction to post-production. The series of change detection images on the right hand side show the difference in vegetation before and after the fire that occurred after production ended. The infrastructure of the bridge had not been removed yet when the fire progressed to Ahamnson Ranch on 9-29-05 (fire began on 9-28) . The results of the change detection analysis are below.
Change Detection - Mission Impossible III
Red = increased vegetation
Blue = decreased vegetation
Red = increased vegetation
Blue = decreased vegetation
Mid-Prep & After Production ---------------------------------------------------------------------------------------------------------------------------------Before & After Fire
These images of Ahmanson Ranch are taken from Google Earth. The image on the left was taken on 5-25-05, before the bridge construction. The image on the right was taken on 12-21-05, after the fire.
The overall blueness of the right hand image indicates a decrease in vegetation over all of Southern California between the end of Sept and beginning of Oct. The large dark blue patch on the bottom left of the right hand image is the location of the fire. If you look closely at that dark blue patch you can see the sharp square edge which divides Ventura and Los Angeles Counties.
I created a subset of Ahmanson Ranch from the Southern California image and overlaid it onto Google Earth Pro. The white line separates Ventura County (to left) and LA County (to right). There is a stark contrast between these two change detection images. Within less than a month the vegetation in the region decreased dramatically. The large blue region on the right indicates the extent of the fire.
This is a close up of the production site. The blue areas (decreased vegetation) in the image to the left by the bridge location are potentially linked to the bridge construction. The image on the right shows a conglomerate of red regions (increased vegetation) between the bridge and Crummer Ranch Road. This indicates that the bridge acted as a fire blockade and protected vegetation.Aerial Photography From Helicopter: Filming of MI-III on 9-03-05
Photo Credit: Scott Trimble
Photo Credit: Scott Trimble
Southwest view of base camp, off in distance is bridge
Northwest view of bridge, Spine Flower quarantined in orange netting
The road was paved for the construction of the bridge, Spine Flower sectioned off
Northward view of bridge
Northward view of bridge
Filming of bridge attack scene, car crashes on bridge
Spine Flower quarantined in the middle of base camp
The Google Earth image to the right was taken on 12-21-05. The base camp was located where that large patch of lighter colored dirt is on the right side. An outline of the bridge and road can be seen at the center of the image. I used Google Earth Pro measuring tools to draw a polygon around each site, base camp is 3.77 acres and the bridge construction area is 14 acres. Approximately 1 acre of Spine Flower was fenced off.
LAY OUT OF PRODUCTION
The Google Earth image to the right was taken on 12-21-05. The base camp was located where that large patch of lighter colored dirt is on the right side. An outline of the bridge and road can be seen at the center of the image. I used Google Earth Pro measuring tools to draw a polygon around each site, base camp is 3.77 acres and the bridge construction area is 14 acres. Approximately 1 acre of Spine Flower was fenced off.DISCUSSION
In the case of Mission Impossible III, legalities surrounding the protection of the endangered San Fernando Valley Spine Flower make a change detection analysis very valuable. Since protecting the flower unexpectedly became an issue, knowledge of how the production affected the vegetation would have been useful. The ability to produce scientific data from the production’s side would give the location manager leverage. Without this data, a location manager must rely on the land owner to dictate damages. Change detection helps location managers make informed decisions about location restoration.
Since a fire swept through the production site less than a month after MI3 wrapped filming, the ability to compare how the production affected the land versus how the fire affected the land is optimal with remote sensing analysis. In order to deduce actual real world significance of the change detection results, it would require collaboration between the location manager and landowner.
If the film’s location staff had known about the Spine Flower in advance, they could have avoided the situation all together by using a different location. This is where geospatial information systems (GIS) comes in. Records of the geographical locations of these critical plant and animal species exists, it just comes down to gaining access to the sensitive information.
Proof of concept: I originally wanted to perform a change detection analysis specific to the San Fernando Valley Spine Flower. I would have done this by identifying the spectral signature of the endangered plant species and mapping out the plant’s location by analyzing the infrared and visible wavelengths of the electromagnetic spectrum. Identifying the spectral signature would have required finding the information in a scientific paper or collecting the information with fieldwork. However, the 30 meter spatial resolution of the Landsat TM sensor that I had to work with was too large to identify the tiny flower. If I had access to higher resolution imagery from before and after the production (this exists but costs money), I could have proved this concept. Instead, I decided to focus my research on learning about resources available to location managers to improve their methods of land restoration and conservation.
photo credit: Anuja Parikh and Nathan Gale
Photography of Bridge Construction: 8-17-05
Photo Credit: Scott Trimble
