Identifying uncertainties for quantifying shrub encroachment in Northern Alaska using LiDAR
Abstract
Climatic warming in the tundra biome has occurred at rates two to three times higher than the global average during the last 150 years, and is continuing at an unprecedented pace. Over the past few decades, Arctic tundra ecosystems have responded to climatic warming via shifts in vegetation composition and increased woody biomass. The resulting increases in size, abundance, and range of deciduous shrubs has affected ecosystem processes, causing shifts in carbon (C) and nitrogen (N) pools that could result in both positive and negative climate feedbacks. In Northern Alaska, the expansion of shrubs has increased by roughly 16% in land area since 1950; however, progress towards quantifying the impact of shrub expansion on biogeochemical cycling at regional scales is limited. Satellite remote sensing of the arctic has been proposed by many as a key tool to better quantify change and ecosystem responses to a warming tundra over large spatial extents, but has primarily only been inferred from multi-year trends of the Normalized Difference Vegetation Index (NDVI). These, and other passive remote sensing studies, have been complicated by the high cloud cover, low solar angle, and low levels of sunlight inherent to the arctic. Additionally, many such studies lack necessary ground validation, resulting in high levels of error, uncertainty, and misinterpretation of results. Improved techniques are therefore needed to quantify and scale how the dramatic biophysical and ecological changes occurring in arctic tundra ecosystems will affect C and N pools at local and regional scales. To address these problems, we tested the ability of active remote sensing (terrestrial (TLS) and airborne lidar (ALS)) at multiple spatial resolutions to better quantify aboveground shrub biomass, vertical, and horizontal canopy cover. Uncertainties in the ability of ALS to quantify shrub allometrics was determined using regression analysis with a stepwise selection procedure to determine the ideal sampling density (starting with 30 points per m^-2) that allows for the smallest root mean square error (RMSE) between the predictor (ALS data at different sampling densities) and explanatory (ground biomass, and TLS data) variables. The results from our field tests will then be used to determine the accuracy with which a sampling density of 8 points per m^-2 (collected along the Dalton highway in 2011) can be used to create a latitudinal baseline map of shrub biomass and abundance.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2013
- Bibcode:
- 2013AGUFM.B51H0401M
- Keywords:
-
- 0480 BIOGEOSCIENCES Remote sensing;
- 0315 ATMOSPHERIC COMPOSITION AND STRUCTURE Biosphere/atmosphere interactions;
- 0428 BIOGEOSCIENCES Carbon cycling;
- 0439 BIOGEOSCIENCES Ecosystems;
- structure and dynamics