Seminars & Events

Seminars:

• ECSPrecip Seminar Series 2024: #3
  Precipitation in all its forms
  Holly Mallinson, Kaidi Peng and Fraser King

• ECSPrecip Seminar Series 2024: #2
  AGU Precipitation Student Award 2023 Winners
  Omid Zandi and Xinyue Liu

• ECSPrecip Seminar Series 2024: #1
  AGU Precipitation Student Award 2023 Winners
  Florian Morvais and Kenza Tazi

• ECSPrecip Seminar Series 2023: #3
  Introduction to various precipitation products: Satellites/Field Experiments/Models:
  Dr. Lynn McMurdie, Dr. Joe Turk, Dr. Kerry Emanuel

• ECSPrecip Seminar Series 2023: #2
  AGU Precipitation Student Award 2022 Winners
  Manish Dhasmana, Benjamin Goffin, Dr. Ishrat Dollan

• ECSPrecip Seminar Series 2023: #1
  AGU Precipitation Student Award 2022 Winners
  SeungUk Kim, Mochi Liao, Dr. Julia Shates

• ECSPrecip Seminar Series 2022: #2
  Introduction to various precipitation products: Satellites/Field Experiments/Models:
  Dr. Jackson Tan, Dr. Brenda Dolan, Dr. Elinor Martin

• ECSPrecip Seminar Series 2022: #1
  Alyssa Stansfield, Fraser King and Yoonjin Lee

Events:

• AGU Fall Meeting ECSPrecip Event 2023
  Dr. Richard Roy, Dr. Sarah Ringerud and Dr. Ana Barros

• AGU Fall Meeting ECSPrecip Event 2022
  Dr. Christopher Kidd, Dr. Rachel McCrary and Dr. Yixin 'Berry' Wen

• AGU Fall Meeting ECSPrecip Event 2021
  Dr. Dalia Kirschbaum, Dr. Zeljka Fuchs and Dr. Randy Chase

• AGU Fall Meeting ECSPrecip Event 2020
  Dr. George Huffman, Dr. Ralph Ferraro, Dr. Efi Foufoula-Georgiou, Dr. Giulia Panegrossi and Dr. Mauricio Zambrano-Bigiarini

ECSPrecip Seminar Series 2024: #3

"Precipitation in all its forms"

26 September 2024 - 15 UTC 16 UTC

(8 am PDT - 9 am MDT - 10 am CDT - 11 am EDT)

(9 am PDT - 10 am MDT - 11 am CDT - 12 pm EDT)

Due to a scheduling conflict we moved the seminar 1 hour forward, same day, same link.

Third of the 2024 quarterly seminars organized by the ECSPrecip subcommittee dedicated to different precipitation types. The focus of this webinar is to introduce folks to different types of precipitation, the science that is currently going on around this type of precipitation, and who the scientists are leading the charge. Holly Mallinson (NASA Goddard Space Flight Center, USA), Kaidi Peng (University of Wisconsin-Madison, USA) and Fraser King (University of Michigan, USA) will provide an overview of the current status of precipitation science for the different precipitation types.

Zoom Link to the Seminar

Dr. Holly Mallinson

Dr. Holly Mallinson is a NASA Postdoctoral Program Fellow at the NASA Goddard Space Flight Center in Greenbelt, Maryland.  Her research focuses on better understanding the interactions between convective storm dynamics and microphysics using a combination of data from field campaigns and high-resolution numerical models.  She previously received a BS from the University of Louisiana-Monroe and a MS and PhD from the University of Illinois Urbana-Champaign where her dissertation focused on how anthropogenic climate change may influence hail processes and subsequent surface hailfall.  Outside of research she’s served as the inaugural chair for the American Meteorological Society’s (AMS) Board on Student Affairs and the Board on Representation, Accessibility, Inclusion, and Diversity.  

ABSTRACT: What the Hail?  An Overview of Hail Science and Research

Hail is a rare hydrometeor that occurs in intense convective storms; it is also amongst the largest hydrometeors with diameters ranging from 0.5 cm to over 18 cm.  Hailstones form when a hail embryo, a frozen raindrop or ice particle that serves as the core of the hailstone, encounters supercooled water within storm updrafts.  Once the hailstone falls out of the updraft (or grows so large that the updraft cannot support it), it falls towards the surface and will melt (either partially or fully) as it encounters warm air.  

Many fundamental knowledge gaps, particularly regarding large hailstones, remain due to the intricate balance between dynamic and microphysical processes required for formation and growth.  Further, airborne and ground-based radars have difficulty accurately observing in-storm hail properties.  As a result, the representation of deep convective storms and hail in numerical weather and climate models remains an area of improvement.  

This talk will highlight hail processes as well as current research in the field of hail physics, including an overview of how observational and numerical model datasets can be used in tandem to better understand hailstorms and how anthropogenic climate change may influence changes in hailstorms.  

Kaidi Peng

Kaidi Peng is a Second-year (Third-year since 2024 Fall) PhD student from the Department of Civil & Environmental Engineering, University of Wisconsin–Madison. She is currently working with Prof. Daniel Wright and Dr. Yagmur Derin focusing on satellite precipitation uncertainty quantification and associated effects in hydrologic predictions. She is trying to convert the precipitation uncertainty into decision-making information for flood and other water-related disaster warnings. Her research interest generally lies in leveraging satellite products and physical and statistic models to solve real-world problems, especially in gauge-limited areas. She is also a Future Investigator in NASA Earth and Space Science and Technology since 2024.

ABSTRACT: Global Multiscale Uncertainty Estimation for Satellite Precipitation to Improve Hydrologic Prediction.

Satellites can measure precipitation at high resolution and in near real-time across the globe. The usefulness of these measurements is limited, however, by their relatively large errors, which exhibit complex variability in space and time. Reliable multiscale precipitation uncertainty estimation could mitigate this limitation, but is difficult due to a lack of “ground truth” data from rain gauges in many parts of the world. Furthermore, coupling uncertainty information with ensemble methods could provide a convenient way to propagate precipitation uncertainty into water prediction models to assist in early decision-making. This task is also difficult though, because the complex space-time structure of precipitation errors is hard to characterize and depict as ensembles. In this study, we demonstrate an end-to-end solution to this combination of challenges. We introduce a technique for generating continental-to-global near real-time precipitation ensembles conditioned only on satellite observations, obviating the need for ground truth information. These ensembles remove the long-term systematic error in global multi-sensor precipitation datasets such as IMERG, while modeling the random error associated with the unknown true precipitation rate. From here, our ensembles are further used to drive WRF-Hydro to simulate streamflow ensemble with propagated uncertainty. We will demonstrate the implication of satellite precipitation uncertainty in the application of streamflow prediction.

Dr. Fraser King

Fraser King is a machine learning and remote sensing scientist with a PhD from the University of Waterloo, specializing in snowfall prediction, statistical modeling, and deep learning. He currently works at the University of Michigan as a research fellow, focusing on generative inpainting models for radar applications and unsupervised learning of hydrometeor microphysical observations. His research interests also include enhancing model interpretability to improve the transparency and reliability of AI-driven precipitation models.

ABSTRACT: Decoding Signals in Snowfall Observations

Snow plays a crucial role in Earth's climate and water cycle, influencing weather patterns and water resources. A comprehensive understanding of snowfall particle microphysics is crucial for improving the accuracy of remote sensing snowfall retrievals and model estimates. This study examines snowfall particle size distributions (PSDs) from across the Northern Hemisphere, using Principal Component Analysis (PCA) to identify key, underlying patterns in the data. The PCA revealed three main features that explain 95% of the variability in the data, linked to snowfall intensity, hydrometeor density, and precipitation regime. These findings enhance our understanding of snow processes across various regional climates, aiding in better snowfall prediction in future studies.

ECSPrecip Seminar Series 2024: #2

"AGU Precipitation Student Award 2023 winners"

14 June 2024 - 3pm UTC

Second of the 2024 quarterly seminars organized by the ECSPrecip subcommittee dedicated to the "AGU Precipitation Student Award 2023 winners". Omid Zandi (University of Arizona, USA) and Xinyue Liu (National University of Singapore, Singapore) will be presenting their work.

Watch the recording

Omid Zandi

Omid is pursuing his second year as a PhD student in Hydrology and Water Resources Sciences at the University of Arizona. His research journey began at the University of Tehran during his master's program, where he developed spatial precipitation data for Iran's orographically complex regions. He utilized a combination of gauge data, precipitation products, and related environmental factors, applying spatial non-stationery methods and ensemble machine learning algorithms.

At the University of Arizona, Omid's research expands on this foundation. He is merging spaceborne microwave and infrared observations of the cloud’s microphysical properties and brightness temperature along with meteorological variables, using machine learning and deep learning models to develop long-term precipitation retrievals focusing on high latitudes. Additionally, Omid is keen on adapting these advanced techniques for real-time precipitation forecasting to ensure the prompt availability of precipitation data.

ABSTRACT: Developing a Novel Long-Term ML-based Precipitation Product from AVHRR in High Latitudes Using Various Observational Datasets and Auxiliary Variables from Reanalysis

Accurate long-term precipitation estimation is critical to studying Earth's hydrologic cycle and energy budget, particularly in high latitudes, where the effect of global warming is significant and vital freshwater reserves are stored in icecaps and glaciers. Unfortunately, in situ observation of precipitation occurrence and intensity is very sparse and uncertain and, due to the lack of high-quality geostationary observations, satellite data records are relatively short in these regions. Here, we present a novel Machine Learning (ML) based precipitation retrieval method that uses brightness temperature and cloud properties from the Advanced Very High-Resolution Radiometer (AVHRR) observations via the PATMOS-X product, environmental information from the MERRA-2, and surface type data from AutoSnow as major input features. Precipitation estimates from coincident CloudSat, GMI MW-based precipitation retrievals are used for training and verification. Where needed, ERA5 precipitation estimates were also introduced. The product is developed at 0.1-degree latitude-longitude poleward of latitude 45 degrees N/S to complement the long-term precipitation record offered by geostationary satellite data. Leveraging the continuous and consistent AVHRR observations across multiple platforms over more than four decades, this product provides a comprehensive long-term precipitation data record in high latitudes that will be helpful to the Global Precipitation Climatology Project (GPCP) product as well as other precipitation products that tend to extend their record back in time, beyond that offered by passive microwave products. This presentation introduces the product and provides early results including evaluation and comparison analyses.

Xinyue Liu

Xinyue is a final-year PhD candidate in the Department of Civil and Environmental Engineering at National University of Singapore (NUS). Her research focuses on hydroclimatic issues, particularly the hazard risks linked to real-life impacts. In the early stages of her PhD, Xinyue was intrigued by the climate factors driving locust plagues in the Middle East and North Africa region. To unravel this complex interaction, she established a data-driven framework to extract spatiotemporal dynamic features, quantify locust-climate relationships, and assess the compound risk of locust invasion. This work was published in Science Advances [Link] and highlighted by several major press media, such as The New York Times [Link], Associated Press [Link], and ABC News [Link]. Currently, Xinyue is delving into the study of spatial compound climate extremes, such as the simultaneous occurrence of extreme precipitation across different regions, which can lead to more devastating impacts than the sum of individual events. She developed a knowledge-informed deep generative model that accounts for the spatial dependence structure of compound climate extremes. Xinyue presented this academic work at the Annual Meeting of the American Geophysical Union (AGU) and won the Precipitation Student Award.

ABSTRACT: A Knowledge-Informed Deep Generative Framework for Risk Assessment of Spatially Compound Extremes

Climate change is driving an alarming increase in the frequency and severity of extreme weather events. The rapid succession of multiple hazards across diverse regions can result in impacts far surpassing individual events' cumulative effects. However, the rarity of these compound events poses challenges to accurately assessing their occurrence probabilities and cascading impacts, as they represent only a small fraction of the multi-dimensional probability space, necessitating an extensive dataset for robust analysis. While existing studies, such as weather generators, can produce stochastic realizations of observed climate variables, they only maintain site-independent statistical properties and overlook the spatial extremal dependence structure critical in understanding spatially compound events. To address these limitations, we develop a knowledge-informed deep generative framework, dubbed ExSPATE-GAN (Extremal embedding SPAtio-TEmporal associated Generative Adversarial Network), to enhance the simulation of spatially compound extremes. Our approach incorporates a spatio-temporal coupling metric as embedding within the GAN model to address the spatial extremal dependence structure. The produced synthetic samples can be employed in hydrological or climate models to anticipate the responses of intricate and nonlinear systems. Such risk assessment will empower policymakers with invaluable knowledge and tools to devise effective strategies and prepare for rare events that can profoundly impact local and regional communities.

ECSPrecip Seminar Series 2024: #1

"AGU Precipitation Student Award 2023 winners"

28 March 2024 - 1pm EDT (17.00 UTC)

First of the 2024 quarterly seminars organized by the ECSPrecip subcommittee dedicated to the "AGU Precipitation Student Award 2023 winners". Florian Morvais (Texas A&M University - Corpus Christi, USA) and Kenza Tazi (University of Cambridge, UK) are two PhD students working on Thunderstorms' microphysics and extreme precipitation.

Watch the recording

Florian Morvais

I studied Physics (License + Master degree) at the University of Reims, France from 2014'ish to 2019'ish. If needed, my two years of Master's degree where focused around "Nanosciences, Optic and Atmosphere" for the 1st, and "Physic, Spectrometry, Engineering and Instrumentation" for the 2nd.

I grew bored of Physics (I wasn't that good at it + I'm not really good at maths) so I did a 6-month internship at the Laboratoire d'Aerologie in Toulouse, France with the subject "Study of lightning thanks to spatial instrument GLM and ground network LMA, from lightning to thunderstorm scale" with Dr. Defer Eric. Where I basically learnt about atmospheric sciences & lightning, dealing with large data, collocation, all that cool stuff that I am now doing during my PhD.

For my PhD, I am working with Dr. Liu Chuntao at Texas A&M University - Corpus Christi. I'm just starting my 3rd year, I don't have a clear subject yet, but I'm working on lightning, passive-microwaves, radar, and a little bit of AI.

ABSTRACT: Comparison of Thunderstorms' Microphysics between the Amazon and Central Africa

The Amazonian rain forest is known to produce maritime-like convective systems; on the other hand, the Congo Basin is known as a lightning hotspot on Earth. This study uses the Precipitation Features (PFs) datasets based on the Tropical Rainfall Measuring Mission (TRMM) and Global Precipitation Mission Core Satellites (GPM) to explore the differences in the microphysics properties of ice inferred from passive microwave and radar observations from space over these two regions. Compared to Congo Basin (CB), the Amazon (AM) produces systems principally composed of smaller ice crystals. For instance, for a PF with similar Ice Water Content (IWC) and/or area of 30 dBZ at -10ºC (Acharge), PFs over the Amazon produce down to half as many lightning flashes compared to PFs in the Congo Basin. A PF with similar maximum 30/40 dBZ echo top heights tends to have lower Brightness Temperatures (TBs) in the 85/89, 165, and 183 GHz frequencies over the Amazon than over the Congo Basin. This suggests that PFs over the Amazon tends to be built around smaller ice particles than those over Congo Basin. At 37 GHz, the Congo Basin tends to have colder brightness temperature than the Amazon, likely due to large hail particles. Using PFs over the Amazon as a reference, a global TBs comparison shows that the median TB value for AM PFs is significantly higher than it is in most oceanic areas but is comparable to areas correlated with higher oceanic lightning activity (e.g., around the South Pacific Convergence Zone). It brings back the idea that although AM PFs during the wet season are similar to maritime clouds, some properties relative to land systems remain.

Kenza Tazi

I am a final year PhD student at the University of Cambridge and the British Antarctic Survey as well as a member of the ‘AI for Environmental Risk’ doctoral programme. My thesis focuses on applying novel probabilistic machine learning methods to better understand and predict precipitation over High Mountain Asia. Outside of my doctoral studies, I have developed machine learning models for other environmental applications such as cloud identification and wildfire forecasting. I am also interested in climate policy and bridging the gap between scientific knowledge and decision-making. Before moving to Cambridge in 2019, I completed an integrated master's at Imperial College London in Physics.

ABSTRACT: Extreme precipitation over High Mountain Asia: assessing likelihoods under different climate scenarios using Bayesian Committee Machines

High Mountain Asia supplies freshwater to nearly two billion via Asia's largest rivers. In this area, rain and snowfall drive river flow. However, future precipitation is still poorly understood in High Mountain Asia. For communities to resiliently adapt to climate change, a more nuanced understanding of future precipitation is needed. To address this challenge, we apply an ensembling method to combine different regional climate model outputs and produce principled uncertainty estimates of precipitation under different climate scenarios. More specifically, Bayesian Committee Machines are fit to the projections of Coordinated Regional Downscaling Experiment (CORDEX) members for the South Asia domain. The results of this study will inform the likelihood of extreme events that could lead to flooding, landslides, or droughts.

AGU Fall Meeting ECSPrecip Event 2023

Three excellent speakers in different career stages of the precipitation field representing industry, academia, and research lab shared their experiences throughout their careers. 

Dr. Richard Roy

Dr. Sarah Ringerud

Dr. Ana Barros

ECSPrecip Seminar Series 2023: #3

"Introduction to various precipitation products:
Satellites/Field Experiments/Models"

14 September 2023 - 3pm EDT

Watch the recording

Download slides:
Satellite Products (Dr. Joe Turk)
Field Experiments (Dr. Lynn McMurdie)
Models (Dr. Kerry Emanuel)

Dr. Lynn McMurdie

Dr. Lynn McMurdie is a Research Professor at the University of Washington Department of Atmospheric Sciences. She has participated in or led several recent field campaigns aimed at understanding precipitation processes.

Knowledge of precipitation distribution, intensity, and patterns of variability is key to the understanding of the global water cycle. Precipitation occurs across a range of scales -- from the local, where individual clouds can produce intense rainfall, to the global, where large-scale circulation patterns produce broad regions precipitation. Daily measurements from rain gauge networks are insufficient to document this range of scales of precipitation variability. In order to fill the knowledge and measurement gaps, many different field campaigns have taken place over the past 50 years, each focused either on a particular region of interest or on particular processes that produce precipitation or both. These field campaigns often include intensive periods where multiple instrument suites, such as radars, lidars, gauges, aircraft-based microphysical instruments and more, are deployed. This presentation will highlight recent field campaigns, including the types of instrumentation used, how to interpret different data types, and how to obtain the data for use in individual research projects.

Dr. Joe Turk

Dr. Joe Turk is a member of the Radar Science and Engineering section at the Jet Propulsion Laboratory (JPL). His overall primary interest is atmospheric passive/active remote sensing related to clouds and precipitation, and the observations needed to improve their representation in weather and climate models. Prior to JPL, he worked for the Naval Research Laboratory’s Marine Meteorology Division. He has been active in NASA’s TRMM and GPM science teams, served as an editor for the AMS Journal of Hydrometeorology, and active in the organization of the CGMS/WMO International Precipitation Working Group since inception. 

Since 1997, satellite-based observations of precipitation have been continuously provided by the data being returned from the NASA/JAXA Precipitation Measurement Missions (PMM), namely the Tropical Rainfall Measuring Mission (TRMM) and the current Global Precipitation Measurement (GPM).  These sensors provide a high level of detail on the horizontal and vertical structure of the precipitation sensed by the GPM radar and radiometer.  However, these sensors provide a more limited description of the environment and surface characteristics.  I will briefly touch on several topic areas in the area of satellite-based measurement of precipitation that remain challenging issues for further investigation.

Dr. Kerry Emanuel

Dr. Kerry Emanuel is an emeritus professor of atmospheric science at the Massachusetts Institute of Technology, specializing in the physics of the tropical atmosphere, including tropical cyclones.

Torrential rainfall from tropical cyclones is a leading case of flooding along and near the U.S. East and Gulf coasts. Yet rain gauge- and radar-based rainfall estimates are problematic in strong windstorms, these records do not span enough time to capture the important 100- and 250-year events, and in any case climate change has already rendered obsolete historically based estimates of flooding hazards. To address this problem, we have downscaled 6,200 U.S.-landfalling tropical cyclone events from each of 8 CMIP6 climate models in each of two climate states. We will give students access to this data set along with software to calculate rainfall return periods at any set of geographical locations. These rainfall events can then be used as input to hydrological flood models.

ECSPrecip Seminar Series 2023: #2

"AGU Precipitation Student Award 2022 winners"

23 June 2023 - 2pm EDT

Second of the 2023 quarterly seminars organized by the ECSPrecip subcommittee dedicated to the "AGU Precipitation Student Award 2022 winners". Manish Dhasmana (IIT Bombay, India) and Benjamin Goffin (University of Virginia) are two research scholars working on extreme flooding events and remote sensing of terrestrial variables for water management information. Ishrat Dollan (Stony Brook University) is a post-doc studying spatiotemporal patterns of extreme precipitation.

Watch the recording

Manish Dhasmana

Manish is a Ph.D. research scholar in the Interdisciplinary Program in Climate Studies (IDPCS), IIT Bombay, India. He is working on the probabilistic event attribution of extreme flooding events on a regional scale. His research aims at carrying out an attribution analysis for hydroclimatic extreme events in India, with a focus on flooding.

ABSTRACT: Evaluation of CMIP6 Models for Extreme Precipitation over India

Climate models are the most important tools available for investigating the response of the climate system to various forcings, for making climate predictions on seasonal to decadal time scales, and for making projections of future climate over the coming century. It is very important to evaluate these models, both individually and collectively. In this study, historical precipitation simulated by the General Circulation Models (GCMs) of the Coupled Model Intercomparison Project Phase 6 (CMIP6) is evaluated against observed gridded data provided by the India Meteorological Department (IMD). Precipitation indices defined by the Expert Team on Climate Change Detection and Indices (ETCCDI) are used for such evaluation. Both moderate and extreme mean precipitation intensities are found to be underestimated by the CMIP6 models across India. Most models overestimate the mean and variability of wet spell durations, and those of dry spells are typically underestimated. The performance of a dataset in reproducing the climatological mean of the reference (IMD) indices is evaluated using root-mean-squared-error (RMSE), and the relative performance of datasets in simulating the inter- annual variability of an index is evaluated by computing the normalized IVSS metric. We also observed that models having a finer spatial resolution, such as the EC-Earth3, MIROC6 and IITM-ESM, perform well with respect to other selected CMIP6 models. This study provides guidance on the selection of CMIP6 models for understanding and assessing future and past changes in precipitation in India and associated impacts.

Benjamin Goffin

Benjamin D. Goffin previously earned a Bachelor of Science in Civil Engineering Technology from Bluefield State University and a Master of Science in Civil Engineering from the University of Virginia. After working in industry for several years, Benjamin returned to the University of Virginia to begin his doctoral research. Funded through a Jefferson Scholars Foundation Fellowship and recently part of the NASA DEVELOP National Program, his current work aims to advance the remote sensing of terrestrial variables and provide decision-relevant information on water processes that are not available by other means. 

ABSTRACT: Changes in the Annual Distribution of Daily Precipitation across the Meuse River Basin  

The catastrophic floods that occurred in Eastern Belgium on July 14-15, 2021, claimed the lives of 39 people and induced heavy property damage (including 50,000 vehicles). One cause of the flooding was the extreme precipitation over the region, with the equivalent of two months of rainfall (over 200 mm in certain locations) occurring in the span of two days. The present work focused on the Meuse watershed, which stretches over 35,000 sq-km from France across Belgium to the Netherlands, and was hard hit by this extreme precipitation event. For this river basin, we analyzed daily Multi-Source Weighted-Ensemble Precipitation (MSWEP) from 1979 to present with a 0.1-degree spatial resolution. We implemented a pixel-by-pixel approach and examined the long-term changes in precipitation variables. We observed substantial increases in the occurrence of storms and in the magnitude of extreme events for the period of 2000-2020 when compared to that of 1979-1999. Additionally, we found robust, significant trends in the annual contribution of the wettest days. Throughout this work, we leveraged available precipitation datasets to characterize spatiotemporal patterns in precipitation across the Meuse River Basin. Overall, this comprehensive study provides decision-relevant facts that can enable better policymaking, appropriate flood mitigation, and further community resilience.

Dr. Ishrat Dollan

Ishrat Dollan earned her MS and PhD in Water Resources Engineering from George Mason University. Ishrat’s doctoral research focuses on understanding spatiotemporal patterns of extreme precipitation across different scales, ranging from regional to continental. Ishrat is going to start her next role as a postdoctoral researcher in the School of Marine and Atmospheric Sciences of Stony Brook University. 

ABSTRACT: How High Mountain Asia’s Precipitation is Changing in a Warming World?

The Himalayan region’s climate and topographic heterogeneity make it challenging to study and predict precipitation and its patterns. Due to the scarcity of reliable ground stations and precipitation dynamics dominated by monsoon (Jun-Sept) and westerlies (Dec-Feb), high-resolution gridded products are fundamental to identifying changes in average and extreme precipitation. In the pursuit of a reliable product over the diverse climatic zone, this work adopts a recently developed ensemble product (LPM, 5km/hourly, 1990-2018) that builds upon the consensus of the existing products from multiple sources (satellite estimates and model reanalysis) over High Mountain Asia (HMA, 20- 46N and 60-95E), rather relying on one single product. Results show significant summer decreasing trends (0.1 alpha) in the western and southeastern and increasing trends in the southern flank of the Himalayas and northwest India. Drying patterns are prevalent in the western Himalayas, in contrast to higher wetting trends over the Tibetan Plateau’s elevated land. Winter increasing trends dominate the western Himalayas and significant decreasing patterns in the eastern Himalayas (east-west gradient). The characteristics of the summer precipitation trends are very localized, with fewer statistically significant regions with increasing trends than decreasing ones in the west and east of the domain. To better understand long-term future extreme precipitation (top 1% exceeding climatological 99th percentile) and change over the complex topography, the study also incorporates bias-corrected (reference product, LPM, 1990-2014) and downscaled (5km) 30 ensembles of Seamless System for Prediction and EArth System Research (SPEAR_MED, 50km) under two socioeconomic narratives, Shared Socioeconomic Pathway (SSPs), ‘middle of the road’ (SSP2-4.5) and ‘fossil-fueled development’ (SSP5-8.5). Results show that both ensemble averages from the SSPs exhibit nearly identical changes until 2060, after which (towards the end of the 21st century) they diverge while showing upward trends. The result indicates that HMA will receive more precipitation from the top 1% in the coming decades under SSP2-4.5 and SSP5-8.5. A comprehensive understanding of historical and projected changes of extreme precipitation under different emission scenarios is imperative to understand future water needs for the HMA region and improve adaptation and mitigation strategies.

ECSPrecip Seminar Series 2023: #1

"AGU Precipitation Student Award 2022 winners"

28 March 2023 - 10am EDT

First of the 2023 quarterly seminars organized by the ECSPrecip subcommittee dedicated to the "AGU Precipitation Student Award 2022 winners". Julia Shates (University of Wisconsin-Madison), Mochi Liao (Duke University) and SeungUk Kim (University of Illinois-Urbana Champaign) are three PhD candidates working on precipitation, investigating rain-snow transition levels in the atmosphere, uncertainties on precipitation estimates in complex terrain and warm season extreme flood events.

Watch the recording

SeungUk Kim

SeungUk is a PhD student in the Department of Atmospheric Sciences at the University of Illinois, Urbana Champaign, where he is a member of the Hydrometeorology Group led by Professor Francina Dominguez.

His research interests include the development of extreme floods and droughts in the US Midwest and the role of land-atmosphere interaction during such events. SeungUk recently completed an analysis of extreme floods, investigating sources of moisture and physical mechanisms, and is currently executing a similar analysis of extreme droughts.

SeungUk earned his Bachelors of Science degree in Atmospheric Sciences and Computational Sciences from Seoul National University in South Korea in 2016. He then served as a weather officer and chief forecaster for the Republic of Korea Air Force at the Seoul Air Base until May 2019.

ABSTRACT: Warm Season Extreme Flood Events in the Midwestern US – Sources of Moisture and Physical Mechanisms 

A comprehensive picture of the development of warm season extreme floods in the Midwestern US is presented. We first identify the climatological moisture sources for precipitation in the Midwest using the two-layer dynamic recycling model (2L-DRM) with ECMWF Reanalysis v5 (ERA5) data. Terrestrial sources supply most of the moisture for Midwestern US precipitation during the warm season, while oceanic sources dominate during the cold season. EOF analysis is used to select extreme flood events characterized by both positive soil moisture and precipitation anomalies. During the warm season flood events, moisture coming from oceanic sources increases by more than 45% compared to the climatology. In addition, our results show that moisture can come from remote regions due to sustained anomalous circulation. Low-level circulation anomalies associated with wave trains that traverse the continent enhance moisture contributions from terrestrial sources along narrow paths. However, moisture budget analysis reveals that the primary flood-producing mechanism is the convergence of moisture due to intense circulation anomalies. Moisture advection and thermodynamic terms are responsible for flood termination. Our results suggest that knowledge of antecedent wet soil moisture conditions is unlikely to improve the predictability of flood-producing storms.

Mochi Liao

Ph.D. candidate Mochi Liao was born in Lixian, Hunan, China, spent 4 years in Wuhan for bachelor degree, and has lived in United States after graduation. He is now pursuing his doctoral degree at Duke University and his final dissertation is scheduled in the summer 2023. His graduate studies were supported by NASA Earth and Space Science Fellowship and National Park Services. 

Mr. Liao is currently a visiting scholar at University of Illinois Urbana-Champaign. His primary research interests are in Physical Hydrology and Hydrometeorology with a focus on understanding and characterizing the spatiotemporal error structure of radar-based Quantitative Precipitation Estimation (QPE) in the mountains, particularly flood-producing precipitation events. 

ABSTRACT: Physics-guided AI framework for estimating precipitation uncertainty in complex terrain 

His research features a multidisciplinary background, such as building a physics-guided QPE error prediction system (Liao and Barros, 2022). The big data analytics he used such as machine learning and deep learning are guided by fundamental physics rather than statistical assumptions, and these techniques are quickly gaining popularity in earth sciences. The resulting data products are accessed by governmental agencies, academic institutions and industry sectors, providing tangible societal benefits.

Julia Shates

Julia Shates is a PhD candidate in the Atmospheric and Oceanic Sciences Department at the University of Wisconsin-Madison.  A primary focus of her research is understanding variability and phase transitions within the satellite radar blind zone that have important implications for the detection and quantification of snow and rain. She leverages observations from ground-based multi-instrument suites to examine micro- and macro-physical characteristics of clouds and precipitation.

ABSTRACT: Multi-year Analysis of Rain-Snow Levels at Marquette, Michigan

 As the climate warms, precipitation is changing and rainfall is replacing snowfall in many regions. Precipitation phase and intensity impact snow cover and accumulation, glacier mass balance, flood forecasting, and safety and transportation across the mid- and high- latitude regions. The height above the surface where ice fully melts to rain, here referred to as the rain-snow level (RSL), influences the precipiation reaching the surface. Ground-based and space-based radars can detect melting; however, space-based radars are limited in their ability to capture precipitation near the surface (< 2 km) due to ground-clutter. Also, rain-snow levels are lower in the mid and high latitudes and during the cold season, which makes detecting melting in the satellite radar blind zone difficult. This study uses ground-based data from an instrument suite at the National Weather Service station in Marquette, MI (MQT) to investigate RSLs in stratiform rainfall in the Upper Great Lakes Region. Rain events from January 2014 to April 2020 are examined using a ground-based vertically profiling radar (Micro Rain Radar; MRR), a custom NASA-developed video disdrometer (Precipitation Imaging Package; PIP), and data products from the ECMWF ERA5 and NASA MERRA-2 reanalyses. Results show distinct macro- and micro-physical characteristics between events with shallow RSLs (< 1.8 km above ground level [AGL]) and intermediate RSLs (> 1.8 km AGL). PIP observations reveal differences in the drop size and fall speed distributions and rain rates as a function of RSL. Seasonal differences, such as shallow RSLs in winter, fall, and spring, have implications for successful detection of melting in the satellite radar blind zone and subsequent identification of surface precipitation phase.

AGU Fall Meeting ECSPrecip Event 2022

Three speakers in different career stages of the precipitation field  shared their experiences throughout their career.

Watch the recording

Dr. Christopher Kidd

Dr. Kidd is currently a Research Scientist at the Earth System Science Interdisciplinary Center of the University of Maryland, working at NASA’s Goddard Space Flight Center. He is a Geographer having studied Geography, Physics and Maths at high school, and going on to study Geography at the University of Nottingham (UK). His PhD, on precipitation retrievals over land, is from the University of Bristol (UK). After his PhD he took up a tenured position at the University of Birmingham (UK), teaching remote sensing, applied meteorology and climatology, as well as research focusing on the retrieval of precipitation from multi-spectral, multi-sensor satellite observations, and representation of precipitation from surface observations (both radar and gauges). His current work involves improving precipitation measurements from both operational and research satellite sensors. He is a previous co-chair of the International Precipitation Working Group and co-leads the IPWG smallsat technology group. He is currently co-lead of the Precipitation-Virtual Constellation working group of CEOS, and a member of the National Academies Committee on Radio Frequencies. He has also served on several advisory panels and committees for new satellite missions. He has a total of over 70 peer-reviewed journal articles and book chapters, together with over 200 conference presentations.

Dr. Rachel McCrary

Dr. Yixin 'Berry' Wen

Dr. Yixin ‘Berry’ Wen earned her M.S. degree of Geoinformatics and PhD degree of Meteorology and her from University of Oklahoma. She did her postdoc in Atmospheric Physics And Weather group at NASA/JPL, then she was a research scientist at the Cooperative Institute for Severe and High-Impact Weather Research and Operations (CIWRO) at University of Oklahoma and National Severe Storms Laboratory at NOAA. She is currently an Assistant Professor in the Department of Geography, University of Florida. She is also a member of the User Working Group in NASA/ Goddard Earth Science Data and Information Service Center. Her research interests include radar and satellite precipitation retrievals, ground validation of remote sensing products, applications of the advanced machine learning technologies in earth science, and climate injustice. 

ECSPrecip Seminar Series 2022: #2

Introduction to various precipitation products:
Satellites/Field Experiments/Models

Watch the recording

Download slides:
Satellite Products (Dr. Jackson Tan)
Field Experiments (Dr. Brenda Dolan)
Models (Dr. Elinor Martin)

Dr. Jackson Tan

Jackson Tan is an Associate Research Scientist with the University of Maryland, Baltimore County, working on the GPM Mission and the IMERG precipitation algorithm at NASA Goddard Space Flight Center. His research focus also includes MODIS and ISCCP cloud regimes, the relationship between clouds and precipitation, and the ground validation of satellite precipitation. He received his Ph.D. in Atmospheric Sciences from Monash University in Australia in 2014.

ABSTRACT: An Overview of Satellite Precipitation Products

Satellite precipitation products are currently the only observation-based source of precipitation with global coverage and, with a record that now extends to decades, are often used in a wide range of scientific and societal applications. In this seminar, I will provide an overview of these satellite precipitation products. I will explain how precipitation is estimated from different types of satellite measurements and how different techniques combine the various streams of observations. From this, I will then summarize the different precipitation products that are available, their advantages and limitations, and—in particular—aspects users should watch out for from the perspective of an algorithm developer.

Dr. Brenda Dolan

Dr. Brenda Dolan’s interests include radar analysis and algorithms, observational integration and validation of cloud-resolving models, precipitation processes, and cloud physics. She is an expert in polarimetric radar, having developed one of the first hydrometeor identification algorithms for short-wavelength radars (X- and C-band). Additionally, she has worked with a wide variety of radars around the world including ground, ship, and satellite-based radars from Ka- to S-band. Brenda has participated in over 10 field projects as a radar scientist or mission scientist. She has worked with large datasets including from radar and disdrometer. Finally she works at the interface of cloud resolving models and observations, using them synergistically to improve observational retrievals and better model representation of physical processes. Recently she has collaborated to develop a polarimetric radar forward operator which can be used with cloud resolving models such as WRF and RAMS.

ABSTRACT: Precipitation-focused Field Campaign Data from Precipitation Radars and Disdrometers 

Field observations of precipitation processes and surface rainfall can fill critical gaps in our understanding of the global water cycle and climate. Assets such as mobile radars, disdrometers, and particle probes can be placed in challenging or data sparse locations such as over the oceans or in complex topography to provide high resolution and detailed observations to better our understanding of precipitation processes and variability, as well as constrain model parameterizations. However, such data is often unique and presents its own challenges for usage, including scanning strategies, data formats, quality control, coverage, and continuity. This talk will focus on radar and disdrometer data collected in recent field campaigns over the oceans and in complex terrain, including a broad overview of available datasets. Specifically, this talk will overview the applications, challenges and advantages of data collected by precipitation radars and disdrometers during field experiments which can be leveraged for precipitation research.

Dr. Elinor Martin

Dr. Elinor Martin is an Associate Professor in the School of Meteorology at the University of Oklahoma (OU), and an OU co-PI of the South Central Climate Adaptation Science Center. The research team she leads aims to understand climate variability and change, with a specific focus on precipitation variability across the Americas and tropics.

ABSTRACT: Precipitation from climate models

This presentation will discuss precipitation from climate models, with a focus on models used as part of the Coupled Model Intercomparison Project (CMIP) Phases 5 and 6. We will discuss what model precipitation data looks like as well as ways to access CMIP5 and 6 model output of precipitation and other fields. We will also address some of the limitations of precipitation in climate model systems and briefly mention climate downscaling for precipitation.

ECSPrecip Seminar Series 2022: #1

Watch the recording

Alyssa Stansfield

Alyssa Stansfield graduated in May 2022 with her PhD in Atmospheric Science from Stony Brook University. Her dissertation research utilized both observations and climate models to study how tropical cyclones and their precipitation are impacted by climate change. This September she will start an NSF AGS postdoctoral fellowship at Colorado State University working with Dr. Kristen Rasmussen.

ABSTRACT

The estimated rate at which climate change increases precipitation within tropical cyclones (TCs) has a wide range of uncertainty. Previous studies using various methods have found rates both below and greatly above the theoretical Clausius-Clapeyron rate of ~6-7% per °C of warming. Modeling studies simulate increasing TC intensities with warming sea surface temperatures (SSTs), which may push the precipitation increase above the Clausius-Clapeyron rate. This work explores and compares the relationship between TC precipitation, SST, and storm intensity in idealized aquaplanet climate model simulations, Earth-like high-resolution climate model simulations, and observations. The methodology involves extracting TC precipitation using an automated algorithm, binning TCs by relevant characteristics (i.e., their local-environment SSTs, intensities, and outer sizes), extracting various precipitation metrics from their precipitation fields, and calculating relationships between the precipitation metrics, TC characteristics, and SSTs.

Fraser King

I am a PhD Candidate at the University of Waterloo studying the intersection of remote sensing and machine learning of precipitation. My background is in computer science and statistical modeling with a master’s focus on remote sensing of snowfall.

ABSTRACT

Radar reflectivity profiles of the lower atmosphere are commonly linked to precipitation through empirical power laws, but these relationships are tightly coupled to particle microphysical assumptions that do not generalize well to different regional climates. Here, we develop a robust, highly generalized precipitation retrieval from a deep convolutional neural network (DeepPrecip) to estimate 20-minute average surface precipitation accumulation using near-surface radar data. DeepPrecip displays high retrieval skill and can accurately model precipitation accumulation, with a mean square error 99\% lower, on average, than current methods. DeepPrecip also outperforms less complex machine learning retrieval algorithms, demonstrating the value of deep learning when applied to precipitation retrievals. Predictor importance analyses indicate a combination of both near-surface (below 1 km) and higher-altitude (1.5 - 2 km) radar measurements are the primary features contributing to retrieval accuracy. This research reveals the important role for deep learning in extracting relevant information about precipitation from radar retrievals.

Yoonjin Lee

I got my master’s and PhD degree in atmospheric science at Colorado State University. I am currently a postdoc at Cooperative Institute for Research in the Atmosphere (CIRA). 

ABSTRACT

The High-Resolution Rapid Refresh (HRRR), current operational forecast model in the United States uses latent heating (LH) derived from ground-based radars to initiate convection during short-term forecast. Although radar reflectivity is a good indicator for LH intensity, and it is efficient in initiating convection, ground-based radar data are not available over the ocean or mountainous regions. Data that are available continuously in such regions and that can be readily used in short-term forecasts are from a geostationary satellite. Despite its inherent limitation of seeing only the cloud top, it can still be useful if taking advantage of its high spatial and temporal resolution data. This study presents methods to detect convection and retrieve LH of the convective clouds from GOES-16, and evaluates impacts of using the retrieved LH in convective initialization for precipitation forecast. The results show that LH from GOES-16 significantly improves precipitation forecast, and it has similar impacts in the forecast if the detection is correctly made.

AGU Fall Meeting ECSPrecip Event 2021

Watch the recording

AGU Fall Meeting ECSPrecip Event 2020

Watch the recording