CSDMS 2015 annual meeting poster HamidDashti
Improving ecosystem dynamic models in a semi-arid ecosystem by integrating different sources of remotely sensed data
Nancy Glenn, Boise State University, Idaho, United States.
Lejo Flores, Boise State University, Idaho, United States.
Nayani Ilangakoon, Boise State University, Idaho, United States.
Jessica Mitchell, Appalachian State University, North Carolina, United States.
Matt Masarik, Boise State University, Idaho, United States.
Lucas Spaete, Boise State University, Idaho, United States.
Qingtao Zhou, Boise State University, Idaho, United States.
Aihua Allison, Boise State University, Idaho, United States.
Abstract:
The Western U.S. semi-arid ecosystems constitute more than 10% of global dryland ecosystems. Dryland ecosystems act as terrestrial carbon sinks as well as habitat for many wildlife. However, drylands have experienced significant degradation, and in particular, the sagebrush-steppe of the Western US. In this region, degradation has occurred due to disturbance from fire, climate, and land use. Monitoring these disturbance effects is ideally performed with remote sensing. Yet, these ecosystems are characterized by sparse vegetation, an abundance of targets with high albedo and non-linear spectral mixing. Moreover semi-arid plant reflectance recorded by airborne and satellite-based sensors are often attributed to canopy structure rather individual leaves. In this study our goal is to use vegetation structural parameters estimated from airborne laser scanning (ALS) and image spectroscopy as constraints in ecosystem models to predict ecosystem structure and fluxes in a semi-arid ecosystem. Our study area is Reynolds Creek Experimental Watershed (RCEW) located in southwestern Idaho, Great Basin. Novel fusion techniques will be used to estimate the vegetation structure products from full waveform ALS (Reigl LMS Q 1560 sensor) and hyperspectral (AVIRIS-ng) datasets collected in late summer 2014. Following a standard protocol, an extensive field campaign was conducted in early fall 2014 and vegetation parameters were measured in the field as well as in the lab for biochemical analysis. Moreover data from our previous experiments such as terrestrial laser scanning measurements (Riegl VZ-1000; 2011 and 2012), multispectral images (RapidEye, 2012) and ALS (Leica ALS50II, 2007) will help us with parameter retrieval. Estimated parameters shall be used to constrain the Ecosystem Demography (ED2) model to quantify ecosystem structures and fluxes. Results will be validated with a network of three eddy flux towers. This study will provide a basis for understanding feedback mechanisms related to changing climate and non-native plant invasion and its impact on wildfire. Our future prospect is to approach a data assimilation framework for integration of remote sensing data and ecosystem models.
* Please acknowledge the original contributors when you are using this material. If there are any copyright issues, please let us know and we will respond as soon as possible.
