Figure 6: Daily precipitation (mm/day) from the MSWEP v2.8 dataset for Hurricane Elena (1985), 
shown within a 10° × 10° box centered on the landfall point. This is a zoomed-in view of the red box 
highlighted in Figure 4, providing greater spatial detail of rainfall evolution around landfall.
Beck, Hylke E., et al. “MSWEP v2 Global 3-Hourly 0.1° Precipitation: Methodology and Quantitative 
Assessment.” AMETSOC, American Meteorological Society, 1 Mar. 2019, 
journals.ametsoc.org/view/journals/bams/100/3/bams-d-17-0138.1.xml. 
Ecmwf. ERA5 Hourly Data on Single Levels from 1940 to Present, 1 May 2025, 
doi.org/10.24381/cds.adbb2d47. 
“MSWEP.” GloH2O, 15 Nov. 2024, www.gloh2o.org/mswep/. 
Summary of the Predecessor Rain Event Prior to Hurricane ..., 
cw3e.ucsd.edu/wp-content/uploads/2024/09/30Sep2024_HelenePRE_EventSummary/HelenePRE_EventSu
mmary.pdf. Accessed 1 May 2025. 
Tropical Cyclone Introduction | National Oceanic and Atmospheric Administration, 
www.noaa.gov/jetstream/tropical/tropical-cyclone-introduction. Accessed 1 May 2025. 
“What Is a Predecessor Rain Event (Pre)?” Climate and Agriculture in the Southeast, 
site.extension.uga.edu/climate/2020/10/what-is-a-predecessor-rain-event-pre/. Accessed 1 May 2025.
 
1. Precipitation Timing and Location Vary Across Storms: 
Although most storms produced peak rainfall on the day of 
landfall, many exhibited substantial precipitation 1–2 days 
prior, with some pre-landfall totals exceeding 50% of 
landfall-day rainfall, as seen in Figure 2. The extent and 
intensity of early rainfall within the 10° × 10° box surrounding 
each landfall location varied widely, indicating that some 
storms begin impacting coastal regions well before official 
landfall.
2. Ring-Shaped Precipitation Structures Are Common: Spatial 
plots often reveal a ring of enhanced rainfall approximately 
100–200 km from the storm center, separated from the core by 
a relative minimum or “moat” in precipitation, as seen in 
Figure 1. This recurring structure likely corresponds to outer 
rainbands commonly seen in landfalling tropical cyclones.
3. Evidence of Potential Predecessor Rain Events (PREs): A 
subset of storms showed organized, long-range precipitation 
occurring far ahead of the storm center, indicative of possible 
PRE-like behavior. These patterns suggest interactions between 
tropical moisture and mid-latitude dynamics, though further 
study is needed to confirm causal links.
 
Results
Tropical cyclones (TCs) that impact the U.S. Gulf Coast are 
often associated with notable rainfall prior to the storm's landfall, 
but the structure and timing of this precipitation remain 
underexamined. Using the Multi-Source Weighted-Ensemble 
Precipitation (MSWEP) dataset, we examined the spatiotemporal 
evolution of daily rainfall in several historical Gulf TC events, 
including the following: Allen (1980), Alicia (1983), Chris (1982), 
and Elena (1985). 
In each case, we see a tendency for a crescent-shaped band of 
enhanced precipitation to form ~3 days before landfall, typically 
around 10° distance from where the TC eventually made landfall. 
In most instances, the band enhancement peaked on the landfall 
day and was usually separated from the cyclone by an area of little 
to no precipitation. 
While we are not establishing causation, the consistent 
co-occurrence of this feature suggests a potential relationship 
between the TC's approach to land and the surrounding 
precipitation structure. Better characterization of these features 
may allow for earlier flood preparedness in advance of landfall, 
and might help elucidate the large-scale factors that affect the 
precipitation that occurs prior to the landfall of TCs.
Abstract
1. Storm Identification: We used the IBTrACS v04r01 dataset to 
identify all U.S. landfalling tropical cyclones from 1971–2020. 
Storms were filtered by selecting records with LANDFALL = 0 
and DIST2LAND ≤ 30 km. The first landfall instance per storm 
was retained using unique storm IDs (SID).
2. Landfall Timing & Precipitation File Matching: Each 
storm’s landfall date was converted into YYYYDDD format to 
match daily files from the MSWEP v2.8 dataset, which 
provides global gridded precipitation data at 0.1° resolution. 
Each file contains one day of precipitation totals under the 
precipitation variable.
3. Spatial Precipitation Mapping: For each storm (193 storms 
total), we generated a sequence of six maps of the Continental 
United States showing daily precipitation from five days before 
to the day of landfall. Each plot includes a 10° × 10° red box 
centered on the storm’s landfall coordinates. The percentage 
above each panel represents the amount of rainfall in the box 
relative to landfall-day precipitation, providing insight into 
spatial and temporal precipitation evolution.
4. Histogram Generation: We created histograms comparing 
pre-landfall rainfall to landfall-day rainfall. Each bar shows the 
percentage of landfall-day rainfall that occurred on a given 
prior day. This allowed us to evaluate the frequency and 
magnitude of early precipitation relative to landfall intensity.
5. Azimuthal Totalling of Landfall-Day Precipitation: On the 
landfall day, we computed azimuthal precipitation totals 
centered on each storm’s landfall location. Using the Haversine 
formula, we calculated distances from the center to each grid 
cell and totaled precipitation in 10 km radial bins from 0 to 
10,000 km, producing radial rainfall profiles.
Data and Methodology
Discussion
References
Investigating the Spatiotemporal Evolution of Pre-Landfall 
Rainfall in Gulf Coast Tropical Cyclones
Sanaa Burke, Shuai Wang
College of Earth, Ocean, and Environment, Department of Geography and Spatial Sciences
Landfalling tropical cyclones often begin delivering impactful 
rainfall well before landfall, and their spatial structure plays a 
critical role in how precipitation is distributed. Recognizing 
features like ring-shaped rainfall and early inland moisture delivery 
is essential for improving flood forecasts and advancing 
storm-centered rainfall modeling. Incorporating storm structure 
into rainfall forecasting could lead to earlier warnings and more 
accurate risk assessments.
Conclusion
Figure 1: Daily precipitation [mm/day] associated with Hurricane Allen (1980), derived from the 
MSWEP v2.8 dataset. Each panel shows total precipitation for one day, from five days prior (top left) 
to the landfall day (bottom right). The red dashed box highlights the CONUS-centered landfall 
region. Percentages indicate the fraction of landfall-day rainfall within the landfall region.
Landfalling tropical cyclones often begin delivering impactful 
rainfall well before landfall, and their spatial structure plays a 
critical role in how precipitation is distributed. Figure 2 shows that 
while the day of landfall typically contains the highest rainfall 
totals, many storms deliver a substantial portion of their 
precipitation (sometimes over 50%) in the days prior. This early 
rainfall varies considerably across storms, both in timing and 
magnitude, underscoring the need for better characterization of 
pre-landfall impacts.
Spatial patterns in Figures 1 and 4 reveal the frequent 
development of crescent- or ring-shaped precipitation structures 
around the storm, often peaking at distances of 100–200 km from 
the center. These rings are separated by precipitation minima or 
“moats” and may correspond to outer rainbands or mesoscale 
convective features not always captured in operational forecasts.
Figure 3 provides further evidence of this ring-like behavior via 
azimuthal profiles of landfall-day precipitation. The peaks in total 
rainfall at intermediate distances support the visual findings and 
confirm that intense rainfall is often concentrated away from the 
storm center. Additionally, in Figure 5, a conceptual schematic of a 
Predecessor Rain Event (PRE) highlights one possible mechanism 
for early inland rainfall delivery, showing how tropical moisture 
can interact with mid-latitude systems to produce heavy rain well 
in advance of landfall.
Figure 3: Azimuthal rainfall profiles showing total precipitation (mm) as a function of distance from 
the landfall point (0–10,000 km) for four storms: a) Alicia (1983), b) Allen (1980), c) Chris (1982), 
and d) Elena (1985). Each plot reflects precipitation structure on the day of landfall.
Figure 4: Daily precipitation [mm/day] associated with Hurricane Elena (1985), derived from the 
MSWEP v2.8 dataset. Note the crescent-shaped “wall” of precipitation that begins to form ahead of 
the tropical cyclone as it makes landfall near the Gulf.
Figure 5: A schematic illustrating the atmospheric setup that led to a Predecessor Rain Event (PRE) 
north of Hurricane Helene. Enhanced water vapor on the storm’s eastern flank was transported 
poleward along an atmospheric river-like pathway by low-level winds and upper-level outflow. This 
moisture interacted with a surface frontal boundary, a jet streak, and upper-level trough dynamics, 
producing heavy rainfall well in advance of Helene’s core. 
Source: CW3E Event Summary – Hurricane Helene PRE (2024)
Figure 2: Probability density histograms showing the distribution of precipitation ratios for the five 
days preceding landfall, compared to the landfall day. Each panel illustrates the proportion of rainfall 
on a given day relative to the total on landfall day, normalized across all storms. The sharp peak at 
100% on landfall day reflects the normalization method.