Characterizing uncertainty in Community Land Model version 5 hydrological applications in the United States

Abstract

Land surface models such as the Community Land Model Version 5 (CLM5) are essential tools for simulating the behavior of the terrestrial system. Despite the extensive application of CLM5, limited attention has been paid to the underlying uncertainties associated with its hydrological parameters and how these uncertainties affect water resource applications. To address this long-standing issue, we use five meteorological datasets to conduct a comprehensive hydrological parameter uncertainty characterization of CLM5 over the hydroclimatic gradients of the conterminous United States. Key datasets produced from the uncertainty characterization experiment include: a benchmark dataset of CLM5 default hydrological performance, parameter sensitivities for 28 hydrological metrics, and large-ensemble outputs for CLM5 hydrological predictions. The presented datasets will assist CLM5 calibration and support broad applications, such as evaluating drought and flood vulnerabilities. The datasets can be used to identify the hydroclimatological conditions under which parametric uncertainties demonstrate substantial effects on hydrological predictions and clarify where further investigations are needed to understand how hydrological prediction uncertainties interact with other Earth system processes.

Fig. 1

A schematic view of the CLM5 benchmark hydrological datasets. In step 2, about 1,300 ensemble parameter sets are generated using a Latin Hypercube Sampling method to produce about 1,300 ensemble time series and error metrics. The same ensemble parameters and error metrics are used in step 3 to generate at-site and regional parameter sensitivity scores as well as behavioral sensitive parameters.

Fig. 2

(a) The 464 CAMELS basins and seven clusters defined by the reproducible k-means++ algorithm. (b) CONUS 1/8° grid cells placed into the same seven clusters. White areas indicate that lakes and wetland are removed in clustering.

Fig. 3

Regional mean monthly flow using the NLDAS-2 forcing data in the 7 clusters. The green spread indicates all ~1,300 ensemble members. The red shading indicates the spread for parameter sets that have annual flow bias within 10% of the observed flows. The blue shading indicates the spread for parameter sets that have annual flow bias within 10% of the observed flows and an NSE value of monthly flow above or equal to 0.5.

Fig. 4

(a) The normalized sensitivity score of the 15 hydrological parameters to the annual flow bias metric at each basin in each cluster. (b) Regional normalized sensitivity score to 28 diagnostic error metrics using Cluster 1-Northeast and NLDAS-2 forcing data as an example.