ABSTRACT Studying the dynamics of gross primary production (GPP) in seasonally dry tropical forests is of fundamental importance to understand the carbon dioxide (CO2) balance in this ecosystem, helping mitigate its potential impacts at the regional and global levels. Thus, the objective of this work was to evaluate the accuracy of GPP estimated via remote sensing in the Caatinga biome. A set of observed data retrieved from micrometeorological towers equipped with eddy covariance systems were used to validate remote sensing data. The set was measured in a preserved Caatinga fragment. Remotely sensed GPP data was retrieved from the MOD17A2 version 6.0 product of the Moderate Resolution Imaging Spectroradiometer (MODIS) sensor onboard the Terra satellite. The validation of MOD17A2 GPP estimates was carried out through the comparison with micrometeorological data measured in situ. In the Caatinga site the comparison between the two GPP data types showed a moderate correlation with Pearson’s correlation coefficient (r) = 0.65 and coefficient of determination (R 2) = 0.43 and the product performed better in representing GPP in the Caatinga during the dry season. Results showed that although the MOD17A2 product represents the annual behaviour of GPP, the algorithm could be improved in order to provide GPP information that is more similar to surface measured data over these land covers.

中文翻译：

---

巴西卡廷加生物群落的 MOD17A2 初级生产总值评估

摘要 研究季节性干燥热带森林中初级生产总值 (GPP) 的动态对于了解该生态系统中的二氧化碳 (CO2) 平衡至关重要，有助于减轻其在区域和全球层面的潜在影响。因此，这项工作的目的是评估通过遥感在卡廷加生物群落中估计的 GPP 的准确性。从配备涡流协方差系统的微气象塔中检索到的一组观测数据用于验证遥感数据。该集合是在保存的 Caatinga 片段中测量的。遥感 GPP 数据是从 Terra 卫星上的中分辨率成像光谱仪 (MODIS) 传感器的 MOD17A2 6.0 版产品中检索到的。MOD17A2 GPP 估计的验证是通过与原位测量的微气象数据进行比较来进行的。在 Caatinga 站点，两种 GPP 数据类型之间的比较显示与 Pearson 相关系数 (r) = 0.65 和决定系数 (R 2) = 0.43 之间存在中等相关性，并且该产品在代表 Caatinga 旱季的 GPP 方面表现更好. 结果表明，虽然 MOD17A2 产品代表 GPP 的年度行为，但可以改进算法以提供更类似于这些土地覆盖的地表测量数据的 GPP 信息。43 并且该产品在旱季代表卡廷加地区的 GPP 表现更好。结果表明，虽然 MOD17A2 产品代表 GPP 的年度行为，但可以改进算法以提供更类似于这些土地覆盖的地表测量数据的 GPP 信息。43 并且该产品在旱季代表卡廷加地区的 GPP 表现更好。结果表明，虽然 MOD17A2 产品代表 GPP 的年度行为，但可以改进算法以提供更类似于这些土地覆盖的地表测量数据的 GPP 信息。