IV. Conclusions

DEM differencing and error analysis can be used with any two DEMs which were taken on different dates. The subtraction of DEMs is rather straight forward, but the process of creating masks, changing projections, reclassifying images, and performing the error analysis make the process more difficult. I have provided somewhat step-by-step instructions for someone with little GIS skills to re-create this research and apply the process to other DEMs.  I do strongly suggest taking a GIS course which will help a lot in this analysis. Throughout this research I have had many problems with the projection, determining the best error analysis, and almost everything else involved with the data. Please feel free to use my knowledge so no one has to go through the same issues. I am available at ksdurst@mtu.edu if you have any questions and I will do my best to help. In addition I have included some future research ideas which would further complement the work I have done here.

Future Research Projects:

1. Continuation of analysis to include 2006 data 
2006 DEMs are currently being created and this research can be extended to include this data utilizing the steps provided here. The 2006 data will most likely be in the GTM coordinate system and will need to be adjusted to fit the 1954 and 2001 dataset. Through this analysis the eruption rate can be determined from 2001 to 2006 at Santiaguito and Pacaya. Although there is no 2001 data for Fuego, I believe there is 1954 data; therefore an analysis could be done on the 1954 to 2006 time period.

2. Convert aerial photographs to DEMs
Many aerial photographs are available for Santiaguito, Pacaya, and Fuego starting from the 1940's.  These aerial photographs can be used to create DEMs, allowing analysis over a much larger temporal scale. In addition, maps created by William Rose for Santiaguito and Otoniel Matias for Pacaya can also be used to determine elevation and volume change.

3.  Detailed GPS survey for Santiaguito, Pacaya, and Fuego
As stated previously, usually DEM error is determined by ground control points. These ground control points are not available for this region, at least not in the quantity needed. A GPS survey of the area would need to be conducted to accurately determine the DEM error.  Hundreds of GPS points would need to be collected in each area.