Seasonal drought predictability and forecast skill over China

Abstract

Under a changing environment, seasonal droughts have been exacerbated with devastating impacts. However, the understanding of drought mechanism and predictability is limited. Based on the hindcasts from multiple climate models, the predictability and forecast skill for drought over China are investigated. The 3 month standardized precipitation index is used as the drought index, and the predictability is quantified by using a perfect model assumption. Ensemble hindcasts from multiple climate models are assessed individually, and the grand multimodel ensemble is also evaluated. Drought forecast skill for model ensemble mean is higher than individual ensemble members, and North American Multimodel Ensemble grand ensemble performs the best. Predictability is higher than forecast skill, indicating the room for improving drought forecast. Drought predictability and forecast skill are positively correlated in general, but they vary depending on seasons, regions, and forecast leads. Higher drought predictability and forecast skill are found over regimes where ENSO has significant impact. For the ENSO-affected regimes, both drought predictability and forecast skill in ENSO years are higher than that in neutral years. This study suggests that predictability not only provides a measure for selecting climate models for ensemble drought forecast in ENSO-affected regimes but also serves as an indicator for forecast skill especially when in situ and/or remote sensing measurements for the hindcast verifications are considered unreliable.

An experimental seasonal hydrological forecasting system over the Yellow River basin – Part 2: The added value from climate forecast models

“…The AC is widely used in the hydro-climate forecast evaluations (Becker et al, 2014;Saha et al, 2014;Mo and Lettenmaier, 2014;Ma et al, 2015), and can be regarded as a measure of forecast skill both in space and time. If the AC is used for each grid cell within the Yellow River basin (i.e., there is only a summation over time), it is reduced to the Pearson correlation.…”

“…The NMME leverages considerable research and development activities on coupled model prediction systems carried out at universities and various research laboratories throughout North America (Kirtman et al, 2014). Besides using the NMME hindcasts for hydrological forecasting over the USA, Europe, southern Asia and global major river basins (Mo and Lettenmaier, 2014;Thober et al, 2015;Yuan et al, 2015a;Sikder et al, 2016), the NMME was also used to assess the potential drought predictability over China (Ma et al, 2015). Given that one of the NMME models, the NCEP-CFSv2, has an ensemble with different initialization dates (Saha et al, 2014), the month-1 forecast is called a forecast at a 0.5-month lead, and the month-2 is at a 1.5-month lead, and so on.…”

Understanding and seasonal forecasting of hydrological drought in the Anthropocene

“…Fortunately, the development of ocean-atmosphere-land coupled general circulation models (CGCMs) provides an unprecedented opportunity to transform advances in seasonal forecasting research (Kirtman et al, 2014;Yuan et al, 2015a) into an integrated drought service. Besides meteorological drought forecasts Ma et al, 2015), agricultural drought forecasts with dynamical seasonal climate forecast models have also been widely applied and evaluated (Luo and Wood, 2007;Mo et al, 2012;Sheffield et al, 2014;Yuan et al, 2015b;Thober et al, 2015). However, dynamical forecasting of hydrological drought based on the CGCM-hydrology coupled approach (Yuan et al, 2015a) has received less attention (Trambauer et al, 2015;Sikder et al, 2016), although there are many statistical forecasting studies for low flows.…”

Weighting of NMME temperature and precipitation forecasts across Europe

“…Because of the large volumes of data that are produced within the NMME (Table 1), global-scale studies have focused on the evaluation of model skill at specific lead times Mo and Lettenmaier, 2014), or for specific seasons (Wang, 2014), models (Jia et al, 2015;Saha et al, 2014), or climate quantities (Barnston and Lyon, 2016;Mo and Lyon, 2015). Regional evaluations of NMME forecast skill have focused principally on North America (Infanti and Kirtman, 2016), the United States (Misra and Li, 2014;Roundy et al, 2015;Slater et al, 2017), the southeastern United States , but also China (Ma et al, 2015a(Ma et al, , 2015b, Iran (Shirvani and Landman, 2016) and South Asia (Sikder et al, 2015). Thus, most of the effort of the NMME model skill evaluation has been over the USA, and far less attention has been paid to Europe, with some exceptions, such as Thober et al (2015), who used NMME forecasts as input for the mesoscale hydrologic model (mHM).…”