Ting-Chen Chen
Ph.D. Atmospheric and Oceanic Sciences
Earth and Life Institute, Earth & Climate
Université catholique de Louvain
Place Louis Pasteur 3/L4.03.08
1348 Louvain-la-Neuve, Belgium
Scroll down for more
About Me.
I am a postdoctoral researcher at the Earth and Life Institute, Université catholique de Louvain, Belgium. In this role, I participate in the Horizon Europe project entitled European Eddy RIch Earth System Models (EERIE), in collaboration with the other 16 partner institutions. My main responsibility is to explore ocean eddies' impacts on/interactions with the atmosphere and sea ice. Recent studies have suggested that resolving these processes, which are not explicitly resolved in current CMIP global climate models, could be crucial for portraying more robust global and regional climate and future projections.
Before joining EERIE, my previous research topics include mesoscale convective processes, their dynamical interactions with synoptic-scale weather systems such as Asian monsoon, tropical and extratropical cyclones(ETCs), and extreme precipitation and wind risk analysis in the context of past, present, and future climates. I obtained a Ph.D. in Atmospheric and Oceanic Sciences at McGill University in Montreal, Canada, in 2021. My Ph.D. thesis was centered on improving the understanding of slantwise convection, i.e., the process by which conditional symmetric instability is released, in a baroclinic environment such as the frontal zones within extratropical cyclones.
Ensemble simulations for a case study, Typhoon Megi (2010), whose track exhibited a sudden curvature, leading to an enhanced remote precipitation effect on the northestern Taiwan.
A case study of the 2019 Halloween Storm, to characeterize ETC-induced local extreme precipitation (mm/h) in Montreal (star sign) using the satellite-based IMERG data. Red dot and the red big circle indicate the cyclone center and its 1000-km radius, identified using an auto-matic cyclone tracking algorithm.
Distribution of 1410 historical ETCs tracked in the northeastern Northern Hemisphere during 2001-2020 in terms of normalized precipitation (x-axis) and 10-m wind speed (y-axis) based on ERA5 reanalysis. Colors indicate the ETC-associated total column water vapor (TCWV). Gray contours indicate the ETC-averaged 850-hPa vorticity.
PROFESSIONAL EXPERIENCES & EDUCATION
2024Jan-present
Postdoc
Université catholique de Louvain,
Louvain-la-Neuve, Belgium
In collaboration with Prof. Hugues Goosse
Participating in the Project “European Eddy-Rich Earth-System Models (EERIE)” to work on the ocean mesoscales and their impacts on the sea-ice, atmosphere mean states, and variability in the Southern Hemisphere.
2022Jul-2023Dec
Postdoc
Karlsruhe Institute of Technology,
Karlsruhe, Germany
In collaboration with Prof. Joaquim Pinto
Participated in the Project“ClimXtreme Module A Physics and Processes” to research the intensity and structural changes of extreme ETCs in a warming climate.
Collaborated with other research teams on "serial cyclone clustering", "evaluation on global reanalysis data for ETC-induced extremes", and "wind-precipitation compound events in Germany".
2021Jul-2022Jul
Postdoc
Université du Québec à Montréal,
Montréal, Canada
In collaboration with Prof. Alejandro Di Luca
Characterized ETC-induced extreme precipitation and winds (distribution, duration, intensity, seasonality) over North America.
Produced a historical cyclone hazard catalog for North America.
Developed a wind risk assessment model using machine learning.
2015-2021
Ph.D.,Atmospheric and Oceanic Sciences
McGill University,
Montréal, Canada
Supervisor: Prof. M.K. (Peter) Yau
Cosupervisor: Prof. Daniel Kirshbaum
Thesis title: "Slantwise Convection: Climatology, Numerical Modeling, and Parameterization" (see more in PHD PROJECTS)
2014-2015
Research assistant
Typhoon Dynamics Research Center (TDRC),
Taipei, Taiwan
Researched the tropical cyclone’s remote rainfall effect, the interaction with the Asian monsoon and topography based on an ensemble simulation.
2012-2014
M.S., Atmospheric Sciences
National Taiwan University,
Taipei, Taiwan
Supervisor: Prof. Chun-Chieh Wu
Thesis Title: "Sudden Track Change of Typhoon Megi (2010) and Its Remote Effect on Rainfall over Taiwan -Evaluation of Uncertainty Based on Ensemble Simulations"
2008-2012
B.S., Atmospheric Sciences
National Central University,
Taoyuan, Taiwan
PUBLICATIONS
Submitted, accepted, or in press
Chen, T.-C., and A. Di Luca, 2024: Characteristics of precipitation and wind extremes induced by extratropical cyclones in North America. Submitted to J. Geophys. Res. Atmos.
Xoplaki, E., F. Ellsäßer, K. M. Nissen, J. Grieger, J. G. Pinto, M. Augenstein, T.-C. Chen, H. Feldmann, and et al., 2023: Compound events in Germany in 2018: drivers and case studies, EGUsphere [preprint], https://doi.org/10.5194/egusphere-2023-1460
Published
Chen, T.-C., C. Braun, A. Voigt, and J. G. Pinto, 2024: Changes of intense extratropical cyclone deepening mechanisms in a warmer climate in idealized simulations. J. Climate, 37, 4703-4722. https://doi.org/10.1175/JCLI-D-23-0605.1
Chen, T.-C., Collet, F., and A. Di Luca, 2024: Evaluation of ERA5 precipitation and 10-m wind speed associated with extratropical cyclones using station data over North America. Int. J. Climatol, 44(3), 729–747. https://doi.org/10.1002/joc.8339
Hauser, S., S. Mueller, X. Chen, T.-C. Chen, J. G. Pinto, and C. M. Grams, 2023: The linkage of serial cyclone clustering in Western Europe and weather regimes in the North Atlantic-European region in boreal winter. Geophys. Res. Lett., 50, e2022GL101900. https://doi.org/10.1029/2022GL101900
Chen, T.-C., A. Di Luca, K. Winger, R. Laprise, and J. M. Thériault, 2022: Seasonality of continental extratropical‐cyclone wind speeds over northeastern North America. Geophys. Res. Lett., 49, e2022GL098776. https://doi.org/10.1029/2022GL098776
Chen, T.-C., M. K. Yau, and D. J. Kirshbaum, 2022: A parameterization of slantwise convection in the WRF model. J. Atmos. Sci., 79, 227–245, https://doi.org/10.1175/JAS-D-21-0131.1
Chen, T.-C., M. K. Yau, and D. J. Kirshbaum, 2021: Sensitivities of slantwise convection dynamics to model grid spacing under an idealized framework. Q. J. R. Meteorol. Soc., 1930-1948, https://doi.org/10.1002/qj.4003
Chen, T.-C., M. K. Yau and D. J. Kirshbaum, 2020: Towards the closure of momentum budget analyses in the WRF (v3.8.1) model, Geosci. Model Dev.,13, 1737-1761, https://doi.org/10.5194/gmd-13-1737-2020
Chen, T.-C., M. K. Yau, and D. J. Kirshbaum, 2018: Assessment of conditional symmetric instability from global reanalysis data. J. Atmos. Sci., 75: 2425-2443, https://doi.org/10.1175/JAS-D-17-0221.1
Chen, T.-C. and C. C. Wu, 2016: The remote effect of Typhoon Megi (2010) on the heavy rainfall over northeastern Taiwan. Mon. Wea. Rev., 144, 3109–3131, https://doi.org/10.1175/MWR-D-15-0269.1
Wu, C.-C., S.-G. Chen, S.-C. Lin, T.-H. Yen, and T.-C. Chen, 2013: Uncertainty and predictability of tropical cyclone rainfall based on ensemble simulations of Typhoon Sinlaku (2008). Mon. Wea. Rev., 141, 3517-3538, https://doi.org/10.1175/MWR-D-12-00282.1
PHD PROJECTS
Key findings:
-
The areas over the midlatitude western oceanic boundary currents exhibit the highest likelihood of slantwise convection in the winter hemisphere (Fig. 1).
-
The probability of CSI release/slantwise convection increases for cyclones with a higher intensification rate (30% on average). A cyclone undergoing rapid intensification has a 57% chance of exhibiting slantwise convection (Fig. 2).
Figure 1: The 37-yr-averaged fractional SCAPE residual for four seasons during 1979–2015. The fractional SCAPE residual is defined as the ratio of (SCAPE-CAPE) to (SCAPE). The closer to one this ratio is, (i.e., SCAPE>>CAPE), the more favorable for slantwise convection the area is. If the atmosphere is close to barotropic or the vertical geostrophic wind shear is nearly zero, SCAPE~CAPE and this ratio falls to zero, favoring upright convection.
-
Compared to the nonexplosive cyclone cases, the time evolutions of SCAPE and VRS (indices for the likelihood of occurrence of slantwise convection) within rapidly deepening cyclones exhibit higher values before, and a more significant drop after, the onset of rapid intensification, suggesting the release of symmetric instability might contribute to the intensification of storms (Fig. 3).
Figure 2: Probability of slantwise convection occurrence potential given the existence of cyclones with different intensification rates (r), defined by the decrease of central pressure per hour.
Figure 3: Upper panels show the time evolutions of the averaged central sea surface pressure (black curve) and the standard deviation among all cyclones cases in the Northern Hemisphere during 1979-2008. Bottom panels show SCAPE (black) CAPE (pink), VRS (blue) averaged within a radius of 300 km around the cyclone center and their standard deviations. The left column is for the explosive cyclones and the right is for the non-explosive cyclones.
Given the potential importance of slantwisec onvection on the midlatitude climate, one may ask whether slantwise convection can be reasonably represented in global climate models considering most cumulus parameterization schemes do not consider slantwise convection? What resolution is required to capture its important features and larger-scale feedbacks?
In this study, a series of 2D idealized experiments of pure slantwise convection are performed in an initially statically stable environment (Fig. 4) using the non-hydrostatic Weather Research and Forecasting Model, with the horizontal grid lengths (∆y) varying between 1 to 40 km. The evolution of an isolated slantwise convection for ∆y=10 km is shown in Video 1.
Figure 4: Initial conditions for the idealized experiments: (a) relative humidity (shading; %), geostrophic wind (gray solid; m/s), saturation equivalent potential temperature surfaces (black dashed; K), absolute momentum surfaces (black solid; m s-1) and (b) saturation equivalent geostrophic potential vorticity, (shading; PVU). The yellow contours of 0, 0.2, and 0.4 K indicate the initial potential temperature perturbation. (c) the initial SCAPE (bar; J/kg) at the location where the corresponding absolute momentum surface intersects the ground.
Video 1: Simulated slantwise convection for ∆y=10 km. The top panel shows the vertical velocity (shaded; cm/s) and the transverse circulation (vectors). The mid panel shows the accumulated precipitation (mm) and the lowest panel shows the SCAPE (J/kg; initial value in solid line; the current value in grey bar) at each location.
Key findings:
-
The results show that the features and larger-scale feedbacks of the slantwise convection converge numerically when a cross-band grid length (∆y) of 5 km is reached (Fig. 5).
-
During the early development of slantwise updraft, small-scale conditional instability inevitably forms due to differential advection of saturation equivalent potential temperature. ∆y<=5 km can resolve such embedded upright convective process, which further promotes a faster release of SCAPE.
Figure 5: The time evolution of the (a) 99th percentile vertical velocity of the domain, (b) domain-averaged SCAPE difference from the initial time and (c) domain-averaged precipitation. (d) 48-hour- and domain-averaged zonal momentum flux for experiments with different horizontal grid spacings as indicated in the legend. The area used for these calculations is bounded by y=200 and 1000 km.
The findings from Project 2 suggest that numerical models may not adequately resolve critical properties associated with slantwise convection with a horizontal grid spacing coarser than around 5 km.
Most commonly used cumulus parameterization schemes (including the Kain-Fritsch scheme being tested below) only target upright convection but not slantwise convection. Thus, either these schemes should be adapted to account for slantwise convection and its larger-scale feedbacks, or new schemes for pure slantwise convection should be developed. A few studies (e.g., Lindstrom and Nordeng 1992, Balasubramanian and Yau 1995, Ma 2000) have shown that the inclusion of slantwise convective parameterization in numerical models improves forecasts of precipitation, jet, and/or cyclone intensity, but these schemes have not been commonly employed.
In this study, we implement a modified version of the slantwise convection (SC) parameterization scheme (Ma, 2000) in the WRF model. Its impacts on improving precipitation simulations (with stronger upward motion closer to the finer-resolution simulation) are shown for both idealized (video 2) and real case studies (Fig. 6).
Experiments explained:
CTL: The control run. High-resolution simulation that explicitly resolves slantwise convection (taken as ground truth).
40KF: Coarse simulation with the Kain-Fritsch cumulus parameterization scheme activated.
40KFSC: Same as 40KF but with our developed slantwise convective (SC) parameterization scheme activated additionally.
*Similar experimental names are used for the real-case studies but with an "r" appended to the end.
Video 2: Simulated slantwise convection for CTL: ∆y=2 km with no cumulus/convective parameterization schemes, 40KF: ∆y=40 km with the Kain-Fritsch scheme activated for upright convection, and 40KFSC: same as 40KF but with an additional slantwise (SC) convective parameterization scheme. The top panel shows the vertical velocity (shaded; cm/s) and the transverse circulation (vectors). The mid panel shows the accumulated precipitation (mm) and the lowest panel shows the SCAPE (J/kg; initial value in solid line; the current value in grey bar) at each location.
To test the performance of the SC scheme in a real-case simulation, we chose a banded precipitation event near a cold front over the United Kingdom in October 2000.
Fig. 6 shows that the additional SC scheme (40KFSCr) produces stronger precipitation. Compared to 40KFr, 40KFSCr has a closer-to-one ETS (equitable threat score), indicating a better consistency with CTLr and therefore a higher quantitative precipitation forecast skill.
Figure 6: Accumulated precipitation (shaded; mm) over 12 h starting from 1800 UTC 29 Oct 2000 for (a) CTLr, (b) 40KFr, and (c) 40KFSCr. The star in (a) indicates the maximum of 86.7 mm. (d) The equitable threat score at different prescribed rainfall thresholds for 40KFr (blue) and 40KFSCr (red) averaged over the domain in (b) and (c).
PHD SIDE PROJECT
Budget analysis of a tendency equation is widely utilized in numerical studies to quantify different physical processes in a simulated system. While such analysis is often post-processed when the output is made available, it is well acknowledged that the closure of a budget is difficult to achieve without temporal or spatial averaging.
To facilitate our studies on the dynamics of slantwise convection, we built an inline budget retrieval method in the WRF(v3.8.1) model to obtain an accurate budget analysis for some prognostic variables (momentum, u, v, w, and potential temperature, theta). With this tool, the 99th percentile residual term is always smaller than 0.001% of the 99th percentile tendency term in our idealized slantwise convection study. The potential rise of errors via offline (post-processed) budget analyses is also investigated in this work.
This code is publicitly available at https://github.com/ting-chenCHEN/WRFV3.8.1_inline_budget_retrieval (the version for this study is tagged GMD_submission1).
Figure 7: Inline budget analysis of horizontal momentum, v, with each term extracted directly from the model. In each row, the shading in each subplot from left to right shows the term of tendency, flux-form advection (ADV), horizontal pressure gradient force (PGF), Coriolis force (COR) (white contours indicate the values exceeding the color bar), PGF+COR and residual. The black contours indicate the horizontal velocity v of 2 and 6 m/s (positive and negative values shown in solid and dashed lines, respectively). Each row from top to bottom illustrates the budget analysis at 6, 12, 16 hour, respectively.