Change Detection using 75-year Aerial Photo and Satellite Data Sets, Inexpensive Means to Obtain 6 cm Resolution Data, and Developing Opportunities for Community-oriented Remote Sensing through Photography

Some governmental research sites have been in existence for as many as 100 years with ground photography used for documentation starting in the early 1900s(e.g., at the USDA Jornada Experimental Range(JER)(783 km2) in south central New Mexico) If ground photography is properly documented when acquired, it can be used as a baseline for later re-photography assessments of change or for providing historical perspective. The JER historical ground photographic data base contains 3148 scenes and is available from a number of photographic archives. For long-term change detection, earth resources satellites (starting in 1972) have been valuable, but their usefulness can be enhanced when combined with data from vertical aerial photography programs begun in the mid 1930s by USDA (and other agency/aerial photo contractors). This historic aerial photography is generally an untapped data resource. Most of these aerial photos have approximately 1-2 m spatial resolution but are scattered in numerous archives around the United States. For the JER, we have been able to locate and obtain 5500 individual aerial photos which have been indispensable in evaluating vegetation change, locating treatment legacies on the landscape, and assessing rangeland treatment effectiveness. All the acquired aerial photos have been digitized and stored in the JER data base. We have also been exploring the use of small Unmanned Aerial Vehicles (UAVs) to provide hyperspatial aerial photos at about 6 cm spatial resolution to complement the 1-2 m resolution aerial photos. The areas covered by these individual UAV photos are only about 215 x 160 m, so in order to cover larger areas we have had to develop our own methods for mosaicing and analysis. In four years of flying the UAV, we have obtained about 15,000 images over JER and other western rangelands. This method is very cost effective because we are able to use an inexpensive 10 megapixel consumer-grade digital camera on the UAV as opposed to more costly aerial mapping cameras or scanners. The extreme high resolution of the images has found great application in rangeland health assessments. Archiving all these photographic products is a necessity so that future researchers can build their own studies on a firm foundation of past observations and analyses. Although we have experienced enormous advances in automated processing and classification of digital remote sensing data since the 1970s, the results are not always better that when we were working with manual interpretation of photographic products. One example is that manual techniques we used for mapping snow cover extent in the 1970-1980s in the Rio Grande basin provided data with information content that may exceed that from today’s automated methods which result in less discrimination of snow cover beneath the forest canopy. Additionally, continued utilization of photography opens up the possibility of community-oriented remote sensing which can have application to developing research areas like phenology and snow hydrology where the public can contribute photos as a form of ground truth