Font Size: a A A

The Study Of Vegetation Phenology In Central Asia Arid Zone

Posted on:2015-04-03Degree:DoctorType:Dissertation
Country:ChinaCandidate:Y G MaFull Text:PDF
GTID:1220330431492158Subject:Physical geography
Abstract/Summary:PDF Full Text Request
Phenology has been considered as a sensitive and intuitive indicator to reflect ecosystem response to the climate change. Arid region in central Asia is one of largest azonality arid zone and one of the most vulnerable areas of ecosystem and water resources system in the world. Under the background of climate change, Central Asia regional ecological environment problems have seriously restricted the central Asian countries economic and social development, and cause the extensive concern of the international community. The absence of findings about phenological observed data, phenological model construction and relationship between phenology with climate have limited the development of phenology in Central Asia. Conducting spatial-explicit phenology studies in Central Asia will improve our understanding the responses of the terrestrial ecosystems in the region with rapid climate change in recent years.This paper was based the long time-series remote sensing data to analysis the tempral and spatial change of vegetation phenology in last30years in Central Asia. Combined remote sensing phenology data and climate data, New phenological model for typical vegetation was constructed.The future trend of phenology of vegetation was predicted under the different climate scenarios. The main conclusion were:Based on GIMMS (Global Inventory Modeling and Mapping Studies) data from1982to2006and SPOT vegetation long time-series vegetation index(VI) data, the threshold method was selected to calculated the vegetation phenological data in arid zone of Central Asia by the Timesat software. The Mann-Kendall trend analysis and Theil-Sen slope tools were used to assess the spatial-temporal change of three phenological metrics. Combined with land cover data and digital elevation model, varieties of phenological metrics were also analyzed for different vegetation cover types and different elevation zones. According to our results, no significant phenological change was detected for the whole study area, but obvious change was found in some local areas. The area where the start of growth season (SOS) had advanced was also found to have a significant extended length of growth season (LEN). Different vegetation cover types showed remarkably different patterns in phonological changes. The end of growth season (EOS) of Deciduous Broadleaf Forest delayed significantly. Except for the Open Shrub, Closed Shrub and Bare Ground, most vegetation cover types showed an extended LEN. The phenology in different elevation varied obviously. The most advanced SOS, delayed EOS and prolonged LEN detected in the2000-3000zone, which may result from the improvement of water and heat conditions caused by the climate change in this zone.Furthermore, we conducted a new method to establish the essential phenology data and climate data for arid zone vegetation phenology model without the observed phenological data in-situ. Based on findings documented concerned temporal and spatial variation of climate factor, A set of rules for selected "relative purify pixels" around weather stations was presented. Under the restriction of the ruleset, the purify pixels’ climate data could be represented by the observed data from these stations. Based these rules, the purify pixels phenology data were extracted from MODIS250m phenology dataset. Under the support of the Particle Swarm Optimization algorithm (PSO), independent sample datasets of three plant function types, included desert grass, mountain grass and deciduous broadleaf forest were built to fit and access eleven SOS phenology models and four EOS phenology models picked from classical phenology models and latest models about arid zone phenology model. Based the result of fitness and accessment, the affect to different plant function types from temperature, precipitation and photoperiod were discussed. Compared with other models, As a result, the T-P modified model and Alternative model which showed the best performance were proposed as the optimum models. Also,the mountain grass show a lower precision then the grass land,which reason may be the variation of temperature and precipitation in mountains. The result suggest the new method used to establish the phenology is effictive and can be imporve the space match of climate data and phenology data.Based the conclusions documented about the Central Asia climate future change, we hypothesized nine scenarios represent the future climate. The temperature and precipitation data represented the future climate change were imported into the phenology SOS models of desert grass and deciduous broadleaf forest to examine the potential response of vegetation phenology to future climate change. The continuous advance was detected in all scenarios with different range. It is estimated of5-10day advance for desert grass and5-20day for deciduous broadleaf forest.
Keywords/Search Tags:Climate change, Remote sensing, Phenology model, Arid zone, Central Asia
PDF Full Text Request
Related items