Ardö Jonas
Department of Physical Geography and Ecosystem Science, Lund University, Sölvegatan 12, Lund, 223-62 Sweden.
Carbon Balance Manag. 2015 Mar 31;10:8. doi: 10.1186/s13021-015-0018-5. eCollection 2015 Dec.
Africa is an important part of the global carbon cycle. It is also a continent facing potential problems due to increasing resource demand in combination with climate change-induced changes in resource supply. Quantifying the pools and fluxes constituting the terrestrial African carbon cycle is a challenge, because of uncertainties in meteorological driver data, lack of validation data, and potentially uncertain representation of important processes in major ecosystems. In this paper, terrestrial primary production estimates derived from remote sensing and a dynamic vegetation model are compared and quantified for major African land cover types.
Continental gross primary production estimates derived from remote sensing were higher than corresponding estimates derived from a dynamic vegetation model. However, estimates of continental net primary production from remote sensing were lower than corresponding estimates from the dynamic vegetation model. Variation was found among land cover classes, and the largest differences in gross primary production were found in the evergreen broadleaf forest. Average carbon use efficiency (NPP/GPP) was 0.58 for the vegetation model and 0.46 for the remote sensing method. Validation versus in situ data of aboveground net primary production revealed significant positive relationships for both methods. A combination of the remote sensing method with the dynamic vegetation model did not strongly affect this relationship.
Observed significant differences in estimated vegetation productivity may have several causes, including model design and temperature sensitivity. Differences in carbon use efficiency reflect underlying model assumptions. Integrating the realistic process representation of dynamic vegetation models with the high resolution observational strength of remote sensing may support realistic estimation of components of the carbon cycle and enhance resource monitoring, providing suitable validation data is available.
非洲是全球碳循环的重要组成部分。由于资源需求不断增加，再加上气候变化导致资源供应发生变化，非洲大陆面临着潜在问题。由于气象驱动数据存在不确定性、缺乏验证数据以及主要生态系统中重要过程的代表性可能存在不确定性，因此量化构成非洲陆地碳循环的碳库和通量是一项挑战。在本文中，对源自遥感和动态植被模型的非洲主要土地覆盖类型的陆地初级生产估计值进行了比较和量化。
源自遥感的大陆总初级生产估计值高于源自动态植被模型的相应估计值。然而，源自遥感的大陆净初级生产估计值低于动态植被模型的相应估计值。在土地覆盖类别之间发现了差异，常绿阔叶林的总初级生产差异最大。植被模型的平均碳利用效率（NPP/GPP）为0.58，遥感方法为0.46。与地上净初级生产的原位数据进行验证，结果表明两种方法均存在显著的正相关关系。遥感方法与动态植被模型的结合并未强烈影响这种关系。
观察到的估计植被生产力的显著差异可能有多种原因，包括模型设计和温度敏感性。碳利用效率的差异反映了潜在的模型假设。将动态植被模型的现实过程表示与遥感的高分辨率观测优势相结合，可能有助于对碳循环组成部分进行现实估计并加强资源监测，前提是有合适的验证数据。
