Search

Postprocessing forecast precipitation fields from numerical weather prediction models aims to produce ensemble forecasts that are of high quality at each grid cell and, importantly, are spatially structured in an appropriate man-ner. A conventional approach, the gridcell-by-gridcell postprocessing, typically consists of two steps: 1) perform statistical calibration separately at individual grid cells to generate unbiased, skillful, and reliable ensemble forecasts and 2) employ ensemble reordering to link ensemble members of all grid cells according to certain templates to form spatially structured ensemble forecasts. However, ensemble reordering techniques are generally problematic in practical use. For example, the well-known Schaake shuffle is often criticized for not considering real physical atmospheric conditions. In this context, a fundamentally new approach, namely, spatial-mode-based calibration (SMoC), has recently been developed for postprocessing forecast precipitation fields with inbuilt spatial structures, thereby eliminating the need for ensemble reordering. SMoC was tested on 1-day-ahead forecasts of heavy precipitation events and was found to produce ensemble forecasts with appropriate spatial structures. In this paper, we extend SMoC to calibrate forecasts of light and no precipitation events and forecasts at long lead times. We also compare SMoC with the gridcell-by-gridcell postprocessing. Results based on multiple evaluation metrics show that SMoC performs well in calibrating both forecasts of light and no precipitation events and forecasts at long lead times. Compared with the gridcell-by-gridcell postprocessing, SMoC produces ensemble forecasts with similar forecast skill, improved forecast reliability, and clearly better spatial structures. In addition, SMoC is computationally far more efficient.

Spatial information of climatological frequency distribution of daily precipitation is highly valuable for a wide range of applications. Accurate estimation of climatology can be made for gauged locations where quality and lengthy observations are available. For ungauged or poorly gauged locations, however, indirect estimation is needed. One approach is to use a gridded daily precipitation dataset derived from interpolating observations. However, gridded daily precipitation data can be subject to large errors when gauge density is low. In addition, most interpolation methods tend to smooth the extreme values and increase the low ones, leading to unrealistic statistical properties and therefore poor estimation of daily climatology. Another approach is to first derive climatology at gauged locations and then interpolate climatology to ungauged locations. While this approach is likely to be more robust than the first approach, low gauge density can still cause significant errors especially in areas of complex terrain. In this study, we develop a method that postprocesses spatially consistent and rich reanalysis data using accurate observations at gauged locations. At an ungauged location, daily precipitation amounts from the reanalysis are bias-corrected using quantile-mapping guided by frequency distributions of reanalysis data and observations at a nearby gauged location (reference location). The bias-corrected precipitation amounts are then used to estimate the climatology for the ungauged location. This method eliminates the need for interpolation and therefore its adverse effects. Special care is taken in quantile-mapping when extrapolating beyond the range of reanalysis data at the reference location. We evaluate the method at 50 locations in Australia, using the Bureau of Meteorology Atmospheric high-resolution Regional Reanalysis for Australia (BARRA) and precipitation observation network across Australia. These locations are chosen to represent different clima

Water is a limited and highly valuable resource. In many parts of the world, water agencies allocate water according to agreed entitlement systems. The allocations are largely based on water already available in storages and rivers. Water agencies may also issue seasonal water allocation outlooks by anticipating future inflows to the storages and rivers. These outlooks are meant to assist water entitlement holders to plan for their crop planting, irrigation, and participation in water markets. Currently, these outlooks are generally based on historical inflow observations (climatology) and are often determined for a small selection of possible climatic scenarios (e.g., extreme dry, dry, average, and wet). These outlooks have large uncertainties, which require users to manage high risks themselves, leading to inefficient water use. In this study, we investigate the use of ensemble seasonal inflow forecasts to improve the production of seasonal water allocation outlooks through a case study of the Goulburn system in central Victoria, Australia. This is a complex system with active water trade both within the region and outside with the larger connected southern Murray-Darling Basin. In this case study, we integrate Australian Bureau of Meteorology’s seasonal streamflow forecasts with Goulburn-Murray Water’s water allocation to produce fully probabilistic water allocation outlooks. We evaluate the outlooks for three irrigation seasons from 2017 to 2020. We compare these outlooks with those produced from using inflows based on climatology only, an approach akin to the current practice of Goulburn-Murray Water. Using seasonal streamflow forecasts resulted in outlooks up to 60% (average 20%) closer to actual determinations, with uncertainty reduced by up to 65% (average 19%) Improvements were most obvious for short lead times and later in the irrigation season. This is a clear demonstration of how integration of streamflow forecasts can improve end-user products, which ca