Analysis of the displacement error of the WRF–ARW model in predicting tropical cyclone tracks over the Philippines

Abstract

ABSTRACT:The direct positional errors (DPEs) in tropical cyclone (TC) track predictions over the Philippine domain of the Weather Research Forecasting (WRF) model's Advance Research WRF (ARW) was studied. The 3 year dataset (i.e. 2012-2014) was categorized (i.e. tropical storm (TS), severe tropical storm (STS), typhoon (TY)) and the marginal distributions of the DPEs over the short-range (i.e. 84 h) prognosis of the model were analysed. The decreasing magnitudes and spread of DPEs at increasing TC intensity were determined. The median DPEs for the model initialization were 70 km (TS), 40 km (STS), 30 km (TY) which progressively increased by 50-90% (all subsets), 130-220% (TS and TY), 430% (TY) for the 1, 2 and 3 day forecasts, respectively. The variations in timing, frequency and length of significant DPE increments for successive lead time levels, as well as the daily significant DPE increments from the model initialization on the TC subsets were also studied. The formulations of the forecast cone of uncertainty (CONU) were likewise incorporated.

The Effect of the Cordillera Mountain Range on Tropical Cyclone Rainfall in the Northern Philippines

“…We first plotted the TC tracks derived from the Control simulations and compared them with tracks from IBTrACS (black tracks in Figure 3). To evaluate the modelled TC track accuracy, we calculated the direct positional error (DPE) [31] between IBTrACS and the TC tracks from the Control simulations. The DPE is defined as the geodesic distance between TC position in the model simulations and the best track [32].…”

Sensitivity of simulating Typhoon Haiyan (2013) using WRF: the role of cumulus convection, surface flux parameterizations, spectral nudging, and initial and boundary conditions

“…Skamarock et al (2008) provides a more detailed description of the model specifications. PAGASA uses WRF for its operational forecasting over the Philippine Area of Responsibility (Flores, 2019;Aragon and Pura, 2016), and it is also used in studies simulating event-based TC-associated rainfall over the Philippines (Cruz and Narisma, 2016).…”

“…One such LAM is the Weather Research and Forecasting (WRF) Model (Skamarock et al, 2008), developed by the National Center for Atmospheric Research (NCAR), which is used as both numerical weather prediction LAM and regional climate model (RCM). WRF is currently used for operational forecasting in the Philippines by the country's meteorological office -Philippine Atmospheric, Geophysical and Astronomical Services Administration (PAGASA) (Flores, 2019;Aragon and Pura, 2016) -and also used in hindcast simulation and sensitivity studies of TC track and intensity (Spencer et al, 2012;Islam et al, 2015;Lee and Wu, 2018) and associated rainfall (Cruz and Narisma, 2016). It has also been used as an RCM to simulate TC activity in the western North Pacific (WNP) basin and several TCs in the North Atlantic over a 13-year period in a convectionpermitting model under current and future climate conditions (Gutmann et al, 2018).…”

Sensitivity of simulating Typhoon Haiyan (2013) using WRF: the role of cumulus convection, surface flux parameterizations, spectral nudging and initial and boundary conditions

“…Skamarock et al (2008) provides a more detailed description of the model specifications. PAGASA uses WRF for its operational forecasting over the Philippine Area of 130 Responsibility (Flores 2019;Aragon and Pura 2016) and it is also used in studies projecting future changes in temperature and rainfall over the Philippines .…”

“…In consideration of the computational cost in resolving important TC processes, the use of RCMs is a valuable and complementary approach to using 65 GCMs in investigating the potential changes in TCs in the future. One such RCM is the WRF Model (Skamarock et al 2008) which is currently used for operational forecasting by the country's meteorological office -Philippine Atmospheric, Geophysical and Astronomical Services Administration (PAGASA) (Flores 2019;Aragon and Pura 2016) and also used in studies projecting future changes in temperature and rainfall over the Philippines Daron et al, 2018;Tolentino and Bagtasa, 2021). 70 A consensus of past modelling studies of western North Pacific TCs, including TY Haiyan, show the cumulus convection scheme as having the most influence on its intensity over other model parameters such as the planetary boundary layer (PBL) and/or microphysics schemes (Islam et al 2014;Di et al 2019).…”