| タイトル | Recent Efforts to Improve the Near Real Time Forest Disturbance Monitoring Capabilities of the ForWarn System |
| 著者（英） | Spruce, Joseph; Hargrove, William; Gasser, Gerald |
| 著者所属（英） | NASA Stennis Space Center |
| 発行日 | 2013-01-07 |
| 言語 | eng |
| 内容記述 | This presentation discusses the development of anew method for computing NDVI temporal composites from near real time eMODIS data This research is being conducted to improve forest change products used in the ForWarn system for monitoring regional forest disturbances in the United States. ForWarn provides nation-wide NDVI-based forest disturbance detection products that are refreshed every 8 days. Current eMODIS and historical MOD13 24 day NDVI data are used to compute the disturbance detection products. The eMODIS 24 day NDVI data re-aggregated from 7 day NDVI products. The 24 day eMODIS NDVIs are generally cloud free, but do not necessarily use the freshest quality data. To shorten the disturbance detection time, a method has been developed that performs adaptive length/maximum value compositing of eMODIS NDVI, along with cloud and shadow "noise" mitigation. Tests indicate that this method can reduce detection rates by 8-16 days for known recent disturbance events, depending on the cloud frequencies and disturbance type. The noise mitigation in these tests, though imperfect, helped to improve quality of the resulting NDVI and forest change products. |
| NASA分類 | Earth Resources and Remote Sensing |
| レポートNO | SSTI 2220-0245 |
| 権利 | Copyright |
|
