Global gridded precipitation and temperature datasets uncertainty in climate change impact studies

Abstract

Climate change impact studies require a reference climatological dataset providing a baseline period against which to assess future changes and post-process climate model biases. Highresolution gridded precipitation and temperature datasets interpolated from weather stations are available in regions of high-density networks of weather stations, as is the case in most parts of Europe and the United States. In many of the world’s regions, however, the low-density of observational networks renders gauge-based datasets highly uncertain. Satellite, reanalysis and merged products dataset have been used to overcome this deficiency. However, it is not known how much uncertainty the choice of a reference dataset may bring to impact studies. To tackle this issue, this study compares global/near-global precipitation and temperature datasets over 3138 North American catchments (high station-density), and 1145 African catchments (low station-density) to evaluate the dataset uncertainty contribution to the results of climate change studies. These datasets all cover a common 30-year period, so they could all potentially be used as reference datasets for climate change impact studies. The precipitation datasets include two gauged-only products (GPCC, CPC Unified), two satellite products (CHIRPS and PERSIANNCDR) corrected using ground-based observations, four reanalysis products (JRA55, NCEPCFSR, ERA-I, and ERA5) and one gauged, satellite, and reanalysis merged product (MSWEP). The temperature datasets include one gauged-only (CPC Unified) product and two reanalysis (ERA-I and ERA5) products. All combinations of these 9 precipitation and 3 temperature datasets were compared and used as inputs to lumped hydrological models to evaluate their performance. They were also used to evaluate the changes in future streamflows and to assess dataset uncertainty against that of other sources of uncertainty. The climate change impact study used a top-down hydroclimatic modeling chain using 10 CMIP5 GCMs under RCP8.5 and two lumped hydrological models (HMETS and GR4J) to generate future streamflows over the 2071-2100 period. Variance decomposition was performed to compare how much the different uncertainty sources contribute to actual uncertainty. Over North-America, the results showed that all three temperature datasets performed similarly, albeit with the CPC performance being systematically inferior to the other two. Significant differences in performance were, however, observed between the precipitation datasets. The MSWEP dataset performed best, followed by the gauge based, reanalysis and satellite datasets categories, respectively. ERA5 was the best-performing reanalysis, while CHIRPS was the best satellite product. Relative dataset performance was also found to be region-dependent. Results show that gauge-based datasets should be preferred in regions with good weather network density, but that CHIRPS and ERA5 would be good alternatives in data sparse regions. For the climate change impact study over Africa, results show that all combination of precipitation and temperature datasets provide good streamflow simulations over the reference period, but 4 precipitation datasets outperformed the others for most catchments: they are in order MSWEP, CHIRPS, PERSIANN, and ERA5. These best performing datasets differ from the ones identified over North-America, demonstrating the impact of the density of weather stations. For the climate change uncertainty study, the 2-member ensemble of temperature datasets provided negligible levels of uncertainty. However, the ensemble of nine precipitation datasets uncertainty that was equal to or larger than that related to GCMs for most of the streamflow metrics and over most of the catchments. A selection of the best 4 performing reference datasets over Africa (credibility ensemble) significantly reduced the uncertainty attributed to precipitation for most metrics, but still remained the main source of uncertainty for some streamflow metrics. The choice of a reference dataset can therefore be critical to climate change impact studies as apparently small differences between datasets over a common reference period can propagate to generate large amounts of uncertainty in future climate streamflows.

Date	15 Dec 2020
Original language	American English
Awarding Institution	École de technologie supérieure
Supervisor	François Brissette (Supervisor) & Richard Arsenault (Co-supervisor)