辽河三角洲河口湿地典型芦苇群落最大光能转化率模拟

Simulation of maximum light coversion efficiency for a Phragmites salt marsh in the Liaohe River estuarine wetland

CASA模型是研究宏观尺度净初级生产力最常用的模型之一,最大光能转化率是CASA模型的最关键参数,但他难以通过测量和试验获得,只能通过模拟求取。利用CASA模型反演了辽河三角洲河口湿地典型芦苇群落的最大光能转化率,并针对遥感和气象数据的可能误差对最大光能转化率的影响进行了敏感性分析。模拟结果表明:芦苇群落具有极高的碳转化能力,最大光能转化率达1.667g C/MJ,实际转化率达到0.957—1.102g C/MJ。敏感性分析结果显示最大光能转化率模拟值对总辐射和NDVI表现出较强的敏感性,由总辐射误差带来的最大光能转化率相对变化幅度仅为-4.14%—4.56%;模拟结果对NDVI的敏感性随着NDVI的增加而降低,即便是以误差30%考虑,模拟值仍然比较集中在样点的变化范围之内,这些结果表明模拟的芦苇最大光能转化率具有一定的稳定性和可靠性。

Maximum light conversion efficiency （MLE） is a critical parameter for the Carnegie Ames Stanford Approach （CASA） model, which is widely used for modeling net primary productivity （NPP） globally. However, it is difficult to parameterize MLE using experiments and field observation. MLE is fundamental in ecological studies, therefore, modeling MLE is of vital importance and significance. The present study determined the MLE of a Phragmites salt marsh in the Liaohe River estuarine wetland in China. The main objectives of this study were to：（1） determine the MLE of a Phragmites salt marsh; and （2） investigate the sensibility of MLE to environmental factors. Factors included in the CASA model comprised the Absorbed Photosynthetically Active Radiation （APAR） using sunshine duration, which was obtained from the National Meteorological Information Center; Fraction of Absorbed Photosynthetically Active Radiation （FAPAR）, calculated using the Normalized Difference Vegetation Index （NDVI） determined from 16-day Moderate Resolution Imaging Spectroradiometer （MODIS） data; temperature and water stress coefficients, calculated using MODIS reflectance data; and meteorological variables including air temperature, precipitation, sunshine duration, air pressure, water vapor pressure, wind velocity, and relative humidity. Field observations of the NPP of the Phragmites salt marsh were conducted at 54 sampling areas with a size of 250 m×250 m, of which 27 sites were used for modeling, and the remaining for validation. The results showed that the Phragmites salt marsh had a relatively high carbon conversion efficiency, with an average MLE of 1.667 gC/MJ, varying between 1.112 and 2.611 g C/MJ, which was much higher than the proposed value of 0.389 g C/MJ by Potter. It was even higher than that of broad-leaved, coniferous, and theropencedrymion forests. The simulated MLE was sensitive to global solar radiation and NDVI, decreasing with their increase, which was more pronounced at lower values. This indicated that it is important to check data quality and increase the data accuracy of global solar radiation. In the present study, global solar radiation was estimated using sunshine duration, with an accuracy of 95%. The relative range of MLE affected by the error of global solar radiation was from -4.14% to 4.56%. MLE became less sensitive as NDVI increased, whereas the simulated values still fell into the MLE range, but the NDVI error increased by 30%. In practical applications, the differences in MODIS NDVI data were much smaller, suggesting that our results are universal and could be used for other satellite images with different spatial resolutions. The air temperature and precipitation errors had little effect on the simulated results, as MLE was not sensitive to them. The results of the sensitivity analysis increased the reliability and confidence of the simulated MLE for the Phragmites salt marsh, which is of great significance when studying the carbon sink and sequestration potential of Phragmites wetlands in China and other regions globally.