| Autor: |
Pickett-Heaps, C. A., Sunter, P., Cornish, A., Sharples, W., Pegios, M., Laugesen, R., Wilson, C. |
| Předmět: |
|
| Zdroj: |
EA National Conference Publications; 2023, p807-813, 7p |
| Abstrakt: |
The Australian Bureau of Meteorology supports hydrological forecast services at multiple temporal scales - from event-focused forecasts (flood warnings) to days, seasons and decades. These services provide either grid-based (nationwide) or catchment-based (point) forecasts. The Australian Water Outlook (AWO) offers nationwide historical analyses, current status, seasonal forecasts and decadal projections of the Australian surface water balance using The Australian Water Resource Assessment landscape model (AWRA). Catchment forecasting services provide event-based, 7-day and seasonal streamflow forecasts at river gauging stations and major water storages across Australia. The Bureau has embarked on a 10-year strategy to unite forecasting services under a common, physically based modelling system. Consistent with this strategy, the AWO seasonal forecasting service and the catchment-based Seasonal Streamflow Forecasting (SSF) service are being integrated under a single hydrological model. The AWO provides seasonal forecasts of root-zone soil moisture, runoff and actual evapotranspiration with climate forcing from calibrated ACCESSS2 seasonal climate forecasts. The SSF service provides streamflow forecasts at point-locations out to three months ahead. Forecast integration is achieved through statistical post-processing applied to AWO seasonal forecasts to generate seasonal streamflow forecasts. Benefits from forecast integration include full system automation, improved forecast consistency between AWO and SSF, reduced system maintenance, reduced reliance on real-time hydrological observations and improved forecast timeliness. Whilst delivering these benefits, forecast system integration must maintain a comparable level of forecast skill to the existing operational service. The Cumulative Rank Probability Skill-Score (CRPS-S) metric was the primary metric used to assess forecast skill. For each month of the year, approximately 90% of sites indicate forecast skill as either equivalent to or an improvement as compared to a historical climatology. Approximately ~66% of sites indicate a forecast skill as either equivalent to or improved compared to a benchmark model (similar in configuration to the existing operational SSF service). AWO derived streamflow forecasts perform well during high-flow seasons, whereas forecast performance drops during the low-flow seasons in certain regions across Australia. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
