利用MODIS火区辐射功率估算中国田间秸秆燃烧排放的研究

Crop residual burning in field could be very popular in China with vast agricultural area. Its emissions play an important role in atmospheric pollution and climate change. There are limitations in the previous methods in estimating the emissions. In addition to the high uncertainties, using annual statistic data could produce the emission inventories with low temporal and spatial resolution. Another estimation method by using the fire counts or the burned area data from remote sensing could miss most of the crop fires in field in China because of the small sizes and short duration. In this study, we shall develop a new algorithm to estimate the crop burning emissions in China based on the MODIS fire radiative power (FRP) products. First, the most important agricultural zones are selected in China using MODIS fire count data. The survey in the farm family and measurements in field would be performed to obtain the fire duration and crop types. Then, the FRP distribution function would be found to describe the fire variations within each agricultural zone. The fire radiative energy (FRE) could be integrated from FRP distribution function within the fire occurrence at each pixel. The mass of dry burned biomass is derived by relating linearly with FRP values. Finally, the emission inventory would be produced with 1 km and daily resolutions. The efficiency to capture the small crop burning fire would also be examined by comparing with the results in some surveyed counties and in-field observations.

秸秆田间燃烧在中国农村十分普遍，它排放出大量的气态和颗粒态的污染物，进而对气候变化和区域大气环境产生重要影响。以往的估算方法通常需要依赖省级年鉴数据，导致清单的不确定性大、时空分辨率低，或者利用卫星火点资料及过火面积产品，但由于不能反映中国农村田间燃烧面积小、燃烧快的特点而大大地低估了排放。最新研究发现，生物质燃烧量和燃烧过程中释放的辐射能呈现显著的线性关系，但目前研究仅限于对燃烧面积大、持续时间长的森林或灌木大火的研究。如何利用卫星资料准确估算田间小火排放，一直是需要解决的难题。本研究将首先利用MODIS火点资料协助对我国农作物种植区进行划分，并进行野外燃烧的定位和调查，然后，利用MODIS辐射功率数据，获取其随时间分布的概率密度函数。在燃烧持续时段进行积分来计算辐射能，最终达到准确估算我国田间秸秆燃烧排放,建立高时空分辨率的排放清单。该研究有助于进一步展开卫星识别秸秆小火的污染排放。