|کد مقاله||کد نشریه||سال انتشار||مقاله انگلیسی||ترجمه فارسی||نسخه تمام متن|
|4758879||1361712||2018||11 صفحه PDF||ندارد||دانلود کنید|
â¢We applied machine learning to EVI profiles to iden6fy SFD pixels.â¢We analyzed the influence of topographic factors on SFD pixels distribu6on.â¢We analyzed the produc6vity loss of SFD pixels compared to the previous 15 years.â¢Eleva6on proved to be the main topographic driver for SFD pixels distribu6on.â¢A produc6vity loss of 14% occurred compared to the previous 15 years.
In common beech forests the most damaging frosts are those that occur at the end of spring. At that time the fresh new leaves are at a vulnerable stage and risk to be readily killed by the freezing temperatures. The ability to identify late spring frost spatial dynamics is a key issue for understanding forest patterns and processes linked to such extreme event. The aim of this study is to detect, map and quantify the vegetation anomalies that occurred in the mono-specific beech forest of the Lazio, Abruzzo and Molise National Park (Italy) after an exceptional spring frost recorded on the 25th of April 2016. Results showed that, beech forests at lower elevations that had an early greening process were subject to spring frost damage (SFD pixels) and their productivity performance strongly decreased with respect to the previous 15 years; to the contrary the beech forests located at higher elevations did not suffer the spring frost effects (NSFD pixels) thanks to their delayed leaf unfolding phase. The duration of the effects of freezing stress for the SFD pixels was about two months, until the end of June, confirmed by Net Ecosystem Exchange measurements. This greening hiatus led to an average 14% loss of productivity compared to the previous 15 years. Elevation had a significant role on the probability of occurrence of SFD pixels. Productivity loss in SFD pixels was more severe at elevations in the range 1500â1700âm, on steeply terrains and North aspects. This study represents a step forward the systematic use of automated techniques to study areas subject to stress or anomalies from multitemporal satellite imagery and to identify break points and recovery of the greening process.
Journal: Agricultural and Forest Meteorology - Volume 248, 15 January 2018, Pages 240-250