REMOTE SENSING OF BIOPHYSICAL VARIABLES AT MULTIPLE SPATIAL SCALES ALONG A LATITUDINAL GRADIENT IN THE CANADIAN ARCTIC

冻土带生态系统占加拿大陆地表面的很大比例，是环境变化研究中的重要系统。苔原生态系统被认为对气候变化特别敏感，但它们将如何反应仍不清楚。冻土带生态系统模式和过程的变化将通过植被物候以及物种组成和丰度的变化来表现。除了气温上升的明显影响外，还有许多因素控制着加拿大北极的植被生长，包括土壤水分、养分可获得性、土壤类型和地形/微地形，其中一些因素也受到气温变暖的影响。因此，高空间分辨率遥感提供了估计和监测植被变异性的机会，有可能量化控制大面积碳通量的生物物理变量。然而，需要进行详细的现场研究，以校准和验证适当的遥感模型，以估计这些变量。 我的NSERC发现基金的重点将是对加拿大北极纬度梯度(°-75°N)上的多个尺度上的生物物理变量进行建模；提供大约10°C的7月平均温度梯度。尽管对这些纬度地区的生物物理变量进行了研究，但这些研究在很大程度上局限于广泛的空间尺度(即1-8平方公里)，现场支持有限。在加拿大北部，很少有关于生物物理变量与高空间分辨率遥感数据(&lt；10米)之间的联系的研究；也没有关于这些变量如何与生态系统过程(例如，碳通量/净生态系统交换)相联系的研究。我的研究将：(1)在高空间分辨率下量化生物物理变量和光谱反射率之间的关系；(2)模拟生物物理变量和生态系统过程之间的关系；(3)在多个尺度上模拟包括碳交换在内的生物物理变量，以便预测这些变量在空间和时间上的变化。 对生产力建模和估计净生态系统交换有重要意义的变量包括地上植物量、植被覆盖率和吸收的光合作用有效辐射的比例。这些是与许多生态系统过程有关的关键结构和生理变量(例如，植物量在评估碳储量方面发挥着重要作用，是全球变化和生产力模型中的一个重要因素，是影响生物多样性的植物群落结构的一种衡量标准)。这些变量将通过使用近红外摄影和光谱指数的直接测量和照片绘制技术来测量。为了扩大变量的范围，将在实地测量和照片绘制数据之间建立模型，以便与高分辨率(Worldview-2)、中等分辨率(Landsat 8)和粗分辨率(MODIS)遥感数据进行比较。有了这项研究的结果，即在多个尺度上对生物物理变量和生态系统过程进行建模，我们将能够做出与北方发展有关的明智的政策决策。此外，这项研究将帮助我们为生活和运营在加拿大北部的社区和资源行业制定适应战略。最后，对本科生和研究生的培训将有助于下一代北极科学家，特别是在北极实地方法和遥感图像处理方面接受培训的地球和环境科学家。

Tundra ecosystems account for a large proportion of Canada’s land surface and are important systems within the context of environmental change research. Tundra ecosystems are thought to be particularly sensitive to changes in climate, yet it remains unclear as to how they will respond. Changes in tundra ecosystem patterns and processes will be expressed through shifts in vegetation phenology and species composition and abundance. In addition to the obvious impacts of increasing air temperatures, there are a number of factors that serve as controls on vegetation growth in the Canadian Arctic, including soil moisture, nutrient availability, soil type and topography/micro-topography, some of which are also impacted by warming temperatures. Hence, high spatial resolution remote sensing provides an opportunity to estimate and monitor vegetation variability, with the potential to quantify biophysical variables that control carbon fluxes over large areas. However, detailed in situ studies are required for calibration and validation of appropriate remote sensing models to estimate these variables. The focus of my NSERC Discovery Grant will be on modeling biophysical variables at multiple scales along a latitudinal gradient (~64°-75°N) for the Canadian Arctic; providing a mean-July temperature gradient of approximately 10°C. This research will be conducted at Sabine Peninsula (77ºN) and Cape Bounty (75ºN), Melville Island; Boothia Peninsula (71ºN); and Apex River, Baffin Island (64ºN), Nunavut. Although there have been studies examining biophysical variables at these latitudes, they have largely been limited to broad spatial scales (i.e., 1-8 km2) with limited field support. There has been very little research conducted in Canada's North on relating biophysical variables to high spatial resolution remote sensing data (<10 m); nor how these variables are linked to ecosystem processes (e.g., carbon flux/net ecosystem exchange). My research will: (i) quantify the relationships between biophysical variables and spectral reflectance at high spatial resolutions; (ii) model the relationships between biophysical variables and ecosystem processes; and (iii) model biophysical variables, including carbon exchange, at multiple scales in order to project changes in these variables over space and time. Variables of significance for productivity modeling and estimating net ecosystem exchange include above-ground phytomass, percent vegetation cover and fraction of absorbed photosynthetically active radiation. These are critical structural and physiological variables linked to many ecosystem processes (e.g., phytomass plays a significant role in assessing carbon stocks, is an important element in global change and productivity models, and is a measure of vegetation community structure which influences biodiversity). These variables will be measured using direct measurement and photo-plot techniques employing near-infrared photography and spectral indices. In order to scale up variables, models will be developed between field measurements and photo-plot data for comparison to high (WorldView-2), intermediate (Landsat 8) and coarse (MODIS) resolution remote sensing data. With the results of this research, i.e., modeling of biophysical variables and ecosystem processes at multiple scales, we will be able to make knowledgeable policy decisions related to development in the North. In addition, this research will help us develop adaptation strategies for communities and resource industries living and operating in Canada’s North. Finally, training of undergraduate and graduate students will contribute to the next generation of Arctic scientists, specifically earth and environmental scientists trained in arctic field methods and remote sensing image processing.