On the use of MARRMoT for rainfall-runoff modelling in Irish raised bogs

Peatlands cover over 20% of Ireland’s landscape, but most have been disturbed by human activities, including land use changes, which alter their natural hydrological functions. As a result, there is a growing need for restoration measures, which require reliable predictive modelling tools for assessing their feasibility and effectiveness. However, choosing a suitable hydrological model, particularly at the catchment scale, can prove challenging. While simplified conceptual rainfall-runoff models remain indispensable water management tools due to their fewer parameters, less input data, and low computational requirements, a critical issue is the limited number of conceptual models that have been successfully applied to peatlands. This is reflected in the lack of intercomparison studies that explore the performance of different model structures for different peatland types. Here, we report on the use of the Modular Assessment of Rainfall-Runoff Models Toolbox (MARRMoT) to analyze the performance of various model structures for three drained, restored, and natural Irish raised bogs. The framework provides a flexible platform for emulating (to a reasonable extent) and comparing different conceptual models within its structure. We emulated the Wageningen Lowland Runoff Simulator (WALRUS) model, which is designed specifically for lowland catchments, and compared it with the other 47 existing models within the framework. The performance of each model was assessed using four goodness-of-fit (GOF) measures. The results revealed a wide range of applicability, which led to several models being excluded from consideration. While the warm-up and calibration periods were limited to less than one year, the reported GOFs provide an invaluable insight into the dynamic performance of the models and the choice of model structure for simulating surface runoff in Irish raised bogs.