Model-independent constraints on modified gravity models hitherto exist mainly on linear scales[1]. A recently developed formalism presented a consistent parameterisation that is valid on all scales[2]. Using this approach, we perform model-independent modified gravity N-body simulations on all cosmological scales with a time-dependent μ. We present convergence tests of our simulations, and we examine how well existing fitting functions reproduce the non-linear matter power spectrum of the simulations. We find that although there is a significant variation in the accuracy of all of the fitting functions over the parameter space of our simulations, the ReACT[3] framework delivers the most consistent performance for the matter power spectrum. We comment on how this might be improved to the level required for future surveys such as Euclid and the Vera Rubin Telescope (LSST). We also show how to compute weak-lensing observables consistently from the simulated matter power spectra in our approach, and show that ReACT also performs best when fitting the weak-lensing observables. This paves the way for a full model-independent test of modified gravity using all of the data from such upcoming surveys.

迄今为止，对修正重力模型的独立于模型的约束主要存在于线性尺度上[1]。最近开发的一种形式主义提出了一种在所有尺度上都有效的一致参数化[2]。使用这种方法，我们在所有宇宙学尺度上执行与模型无关的修正重力 N 体模拟，并且具有时间相关的 μ。我们展示了模拟的收敛性测试，并检查了现有拟合函数如何很好地再现模拟的非线性物质功率谱。我们发现，尽管在我们模拟的参数空间内所有拟合函数的准确性存在显着差异，但ReACT[3] 框架为物质功率谱提供了最一致的性能。我们评论了如何将其改进到未来调查所需的水平，例如欧几里得和维拉·鲁宾望远镜 (LSST)。我们还展示了如何从我们的方法中模拟的物质功率谱中一致地计算弱透镜可观测值，并表明ReACT在拟合弱透镜可观测值时也表现最佳。这为使用来自此类即将进行的调查的所有数据对修改后的重力进行完全独立于模型的测试铺平了道路。