Establishing an IPWG validation site for precipitation products over India

Chris Kidd, Virendra Singh, A K Mitra, Amit Kumar and S K Mukherjee

ESSIC/UMD and NASA/GSFC, and NMSC India Meteorological Department, New Delhi, India

Abstract

A concerted effort has been made by the IPWG to explore and embrace new validation regions. Over the last twelve months or so a validation region over India has been established using processing codes adapted from the Western European and South African validation sites. The processing of the data is carried out by the Indian Meteorological Department (IMD): surface reference data sets are obtained from rain gauge data available on a daily basis. Satellite precipitation products are generated primarily from hourly products to ensure time matching of the gauge reports and the satellite products, this being accumulated precipitation from 0300Z through 0300Z the following day. Importantly, the Indian validation region is one that encompasses a diverse meteorological and climatological range of regimes, ranging from semi-arid regions in southern India, through to tropical, monsoonal regimes, and to temperature mountainous regimes in the north. An outline of the development of this validation region will be provided, together with some of the initial results, and potential further development of the study region in the future.

Results from IPWG Japan Validation Site

Shoichi Shige, Munehisa K. Yamamoto, Asaka Sugimoto

Kyoto University

Abstract

Microwave radiometer (MWR) rainfall algorithms generally consist of a rain/no-rain classification (RNC) and a rain-rate estimation [brightness temperature (TB)–rain-rate conversion] over the delineated rainy area. The rain-rate estimation has been developed under the paradigm that heavy rainfall is associated with strong cold-rain processes. Therefore, conspicuous underestimation of rainfall by MWR algorithms occurred in coastal mountains of the Asian monsoon region where heavy orographic rainfall is frequently associated with strong warm-rain processes. We have improved the heavy orographic rainfall retrievals in the Global Satellite Mapping of Precipitation (GSMaP) MWR algorithm, incorporating an orographic-nonorographic rainfall classification scheme to identify orographic rainfall associated with strong warm-rain processes enhanced by low-level orographic lifting. However, problems of missing light rainfall in coastal mountains of the Asian monsoon region remain unresolved. While the rain-rate estimation could be improved by considering precipitation processes (e.g. warm-rain vs cold-rain processes), RNCs should be improved considering cloud processes such as the onset of precipitation. In the presentation, improvements of GSMaP MWR algorithm considering cloud and precipitation processes are reviewed.

Evaluation of High Resolution IMERG Satellite Precipitation over the Global Oceans using OceanRAIN

Paul A. Kucera and Christian Klepp

University Corporation for Atmospheric Research and University of Hamburg

Abstract

Precipitation is a key parameter of the essential climate variables in the Earth System that is a key variable in the global water cycle. Observations of precipitation over oceans is relatively sparse. Satellite observations over oceans is the only viable means of measuring the spatially distribution of precipitation. In an effort to improve global precipitation observations, the research community has developed a state of the art precipitation dataset as part of the NASA/JAXA Global Precipitation Measurement (GPM) program. The satellite gridded product that has been developed is called Integrated Multi-satelliE Retrievals for GPM (IMERG), which has a maximum spatial resolution of 0.1º x 0.1º and temporal 30 minute. Even with the advancements in retrievals, there is a need to quantify uncertainty of IMERG precipitation estimates especially over oceans. To address this need, the OceanRAIN dataset has been used to create a comprehensive database to compare IMERG products. The OceanRAIN dataset was created using observations from the ODM-470 optical disdrometer that has been deployed on 12 research vessels worldwide with 6 long-term installations operating in all climatic regions, seasons and ocean basins. More than 6 million data samples have been collected on the OceanRAIN program. These data were matched to IMERG grids for the study period of 20 March 2014-28 February 2017. This evaluation produced over 5000 matched IMERG-OceanRAIN pairs of precipitation observed at the surface. These matched pairs were used to evaluate the performance of IMERG stratified by different latitudinal bands and precipitation regimes. The presentation will provide an overview of the study and summary of evaluation results.

Evaluation of TRMM/GPM Blended Daily Products over Brazil

José Roberto Rozante, Daniel A. Vila, Júlio Barboza Chiquetto, Alex de A. Fernandes, and Débora Souza Alvim

CPTEC/INPE, Dept Geography/USP, CCNH/UFABC

Abstract

The precipitation estimates from the Tropical Rainfall Measuring Mission (TRMM) Multi-satellite Precipitation Analysis (named TMPA and TMPA-RT for the near real-time version) are widely used both in research and in operational forecasting. However, they will be discontinued soon. The products from the Integrated Multi-satellite Retrievals for Global Precipitation Measurement (IMERG) and The Global Satellite Mapping of Precipitation (GSMaP) are analyzed as potential replacements for TMPA products. The objective of this study is to assess whether the IMERG and/or GSMaP products can properly replace TMPA in several regions with different precipitation regimes within Brazil. The validation study was conducted during the period from 1st of April, 2014 to the 28th of February, 2017 (1065 days), using daily accumulated rain gauge stations over Brazil. Six regions were considered for this study: five according to the precipitation regime, plus another one for the entire Brazilian territory. IMERG-Final, TMPA-V7 and GSMaP-Gauges were the selected versions of those algorithms for this validation study, which include a bias adjustment with monthly (IMERG and TMPA) and daily (GSMaP) gauge accumulations, because they are widely used in the user’s community. Results indicate similar behavior for IMERG and TMPA products, showing that they overestimate precipitation, while GSMaP tend to slightly underestimate the amount of rainfall in most of the analyzed regions. The exception is the northeastern coast of Brazil, where all products underestimate the daily rainfall accumulations. For all analyzed regions, GSMaP and IMERG products present a better performance compared to TMPA products; therefore, they could be suitable replacements for the TMPA. This is particularly important for hydrological forecasting in small river basins, since those products present a finer spatial and temporal resolution compared with TMPA.

Validation of TRMM Satellite Data in the Farmington/Du River Basin of Liberia

Eugene V. S. Gar-Glahn1, Omowumi O. Alabi, PhD2

1 Liberia Meteorological Service, Ministry of Transport, Monrovia, Liberia; 2 African Regional Centre for Space Science & Technology Education (ARCSSTE), Ile-Ife, Nigeria

Abstract

This research was focused on the Farmington/Du River Basin (latitude 6.4° North and longitude 10.4° West) located in Margibi county, one of the 15 counties in Liberia. The study compared the Tropical Rainfall Measuring Mission (TRMM) Satellite derived monthly precipitation (TRMM3B42V6) with rain gauge data recorded at one meteorological station in the Farmington/Du River Basin, and the possibility of using satellite estimated rainfall to complement ground-measured values. The monthly rainfall comparisons showed that the TRMM rainfall trends were very similar to the observed data trends. The correlation between the monthly datasets ranged from 0.75 to 0.98. The two sets of data captured the phenomenon known as the ‘little dry season’, characterized by an abnormal decrease in precipitation which occurred during the month of July at the peak of the rainy season. Although the overall catchment rainfall was well represented by TRMM data, it was observed that the annual ground measured values were either overestimated or underestimated. TRMM’s persistent underestimations which occurred from 1998 to 2004 were generally below 18% while the overestimations which were recurrent between 2005 and 2009 were below 9%.This study concluded that even though the TRMM precipitation did not perfectly match with the rain gauge data, it can still be used to supplement ground measurements and for estimating rainfalls in un-gauged basins. This is especially important in Liberia, a country where political instability resulted in the extinction of almost all the few meteorological stations in the region.

Validation of Seasonal Precipitation Properties in Korean Peninsula using GPM and Ground Instruments

Choeng-Lyong Lee, GyuWon Lee, AND Geunhyeok Ryu

Kyungpook National University(KNU), CARE/KNU, and NMSC/KMA

Abstract

The GPM (Global Precipitation Measurements) core satellite is one of efficient instrument for global precipitation mapping. Besides, observation products based on GPM DPR (Dual-frequency precipitation) and GMI (GPM Microwave Imager) allow the analysis of seasonal precipitation over Korean peninsula. The ground validation (GV) of GPM measurements is essential and is still challenging issue. Therefore, we performed the statistical characteristic analysis of GPM validation using long-term period databases from GPM and ground instruments such as ground radar (GR) and gage networks (GN). The GV procedure is consist of comparison for reflectivity and precipitation intensities using time synchronized observations from GPM and ground instruments. To compare directly, both measurements are interpolated in same coordinate system called digital forecast system (DFS). For GR, reflectivity calibration is carried out using self-consistency and inter-comparison. The three dimensional constant altitude plan position indicator (3D-CAPPI) and the Hybrid Surface Rainfall (HSR) are then utilized to compare 3D-reflectivity structure and precipitation intensity, respectively. The daily accumulated precipitation amounts of gauges are also used to calculate the precipitation intensity to apply the GV process. To evaluate the GV result, we investigate the statistical indices consisting of mean bias, root mean square errors (RMSE) and correlation coefficient (CORR). The reflectivity measurements of GPM DPR are 3~5 dB higher than GRs regardless of the seasonal variability. The uncertainties of statistic indices for precipitation decrease in the large spatiotemporal scales. Snowfall observations based on reflectivity of GR (GMI) are larger than those of GPM DPR (GRs) at snowfall intensity above 1 (0.3) mmhr-1

Updating the NOAA AMSR2 Operational Precipitation Algorithm.

Patrick Meyers and Ralph Ferraro

Univ. of Maryland - CICS-MD, NOAA/NESDIS/STAR

Abstract

Rain rate estimates from the Advanced Microwave Scanning Radiometer 2 (AMSR2) are operationally produced by NOAA and distributed to operational forecasting offices nationwide. The Goddard Profiling Algorithm 2010 Version 2 (GPROF2010V2) is used to retrieve global rain rates from AMSR2. GPROF2010V2 meets Joint Polar Satellite System (JPSS) program accuracy requirements, however it has persistent issues separating convective and stratiform precipitation, and overproduces rain over cold semi-arid surfaces. An update to the algorithm, GPROF2010V3, was developed and tested to improve the identification of precipitation over land surfaces by: (1) Using dynamic snow cover observations to replace the climatological snow screening; (2) Eliminating a legacy sea ice screen that was over-flagging retrievals at high latitudes; (3) Updating calibration coefficients for raining scenes; (4) Applying Turk et al. (2016) to reduce widespread false alarms; (5) Updating flagging routines to increase usable data coverage. Additionally, the most recent precipitation algorithm (GPROF2017) developed for NASA’s Global Precipitation Measurement (GPM) mission is being evaluated as a possible replacement for GPROF2010V3. Based on extensive evaluation, GPROF2017 categorically performs better than any version GPROF2010 for AMSR2. GPROF2017 vastly improves rain/no-rain separation and produces more realistic precipitation features. In particular, the algorithm eliminates large regions of false precipitation detection over radiometrically cold surfaces. Prior to delivering GPROF2017 to NOAA/STAR for operational implementation, testing and evaluation must be completed in an offline mode to determine the reliability and performance of the algorithm within NOAA’s operational guidelines. GPROF2017 requires dynamic ancillary data about near-surface temperature, total precipitable water, and snow cover. The near real-time version of GPROF2017 is adapted to use Global Data Assimilation System (GDAS) analyses, and comparisons relative to ground truth demonstrate GPROF2017 is suitable for NOAA operations.

Validation of Satellite Precipitation Estimates over the Continental United States using Unique Rain Gauge Data Sets

Ralph Ferraro, Douglas Miller, Patrick Meyers, Robert J. Kuligowski, Brandon Bush and Jackson Hill

NOAA/NESDIS/STAR, Univ. of North Carolina Asheville, Univ. of Maryland/CICS, NOAA/NESDIS/STAR, Univ. of Maryland/CICS, Univ. of Maryland/CICS

Abstract

Two unique rain gauge data sets are used to validate satellite precipitation estimates over the Continental United States (CONUS) to test their utility for routine product monitoring and also explore different attributes of the satellite products. In this pilot study, we explore the Duke Great Smoky Mountain Rain Gauge Network (Duke GSMRGN) and the Community Collaborative Rain, Hail and Snow Network (CoCoRaHS) to validate 24-hour satellite precipitation estimates from several IPWG-monitored products, including the Self-Calibrating Multivariate Precipitation Retrieval (SCaMPR) and CPC Morphing technique (CMORPH) products. CoCoRAHS is a volunteer network covering much of the CONUS and provides 24-hour totals, typically at 1200 UTC. The GSMRGN is a specialized, very local network in the Smoky Mountains (North Carolina and Tennessee) that covers a wide range of elevations. Its primary purpose is to monitor and study orographic precipitation. This network of about 30 rain gauges requires considerable human intervention to obtain and maintain the gauges, several of which are in rugged terrain and only accessible through hiking trails. Students from the University of North Carolina Asheville are provide much of the data access and sensor maintenance through period visits to the gauges. In this pilot study, selected periods of the SCaMPR and CMORPH products are evaluated through newly developed analysis packages from University of Maryland student interns as part of education and outreach activities through the Cooperative Institute for Climate and Satellites (CICS). Results on the satellite algorithm performance for the pilot study period will be shown in this poster presentation.

Assessing FY-3 MWRI global rain rate products using satellite precipitation radar data

Xiaoqing Li

National Satellite Meteorological Center, China Meteorological Administration

Abstract

Fengyun-3D (FY-3D) polar-orbiting meteorological satellite has been launched on November 11, 2017. Up to now, MWRI global rain rate product (MRR) from FY-3B/3C/3D can be got. The goal of this paper is to assess MRR products from the three FY-3 satellites. Radar can get more believable precipitation intensity data. So precipitation products of Dual-frequency Precipitation Radar (DPR) on board the Global Precipitation Measurement (GPM) core observatory, as reference data, were used to assess FY-3 MRR products in the study. Assessment of Rain rate based on surface type, latitude, and precipitation type was carried out. Except that, rain identification, as one important factor of rain retrieval, was assessed too. The results of precision evaluation of three MRR products were compared in this study.

Verification of Quantitative Precipitation Guidance Forecast (QPGF) Based on Fine-mesh and Rain gauge Stations in China

Dan Qi

Kaicun Wang

Abstract

The fine-mesh quantitative precipitation guidance forecast(QPGF) is to meet the demands of national preventing and mitigating disasters, major social activities, and fine weather services. According to fine-mesh precipitation guidance forecast from May 2017 to August 2017 released by National Meteorological Center(NMC), the fine-mesh QPGF at a resolution of 0.05*0.05 has been verified against two datasets. One dataset is by using the Quantitative Precipitation Estimation(QPE) based on merging multiple data in China, the other dataset is daily rain gauge observation dataset over 50000 spots which released by National Meteorological Information Center (NMIC). The verification has be done against two datasets using Threat Score (TS), Equitable Threat Score (ETS), Peirce Skill Score (PSS) and Confidence intervals (Bootstrap), Mean error (ME), Mean absolute error (MAE), Root mean square error (RMSE) ,etc, which followed standardized verification of NWP products released by WMO. The precipitation was divided into five grades with the consistency of operational standard at the National Meteorological Center (NMC) as no rain, little rain, moderate rain ,heavy rain and torrential rain with thresholds of 0mm,0.1mm (including), 9.9mm (including), 24.9mm (including) and 49.9mm (including) respectively. The results show that in little rain 24h forecasting, the TS score in rain gauge station is higher than that in the fine-mesh spot. False alarm rate and missing alarm rate are similar in both methods. The TS score descends with the rainfall increasing. The probably main reason is that the verification based on fine-mesh is effected by the accuracy of the QPE, which are not as real as it on the rain gauges. Meanwhile, the inspection methods contributed to this result.

Rainfall observation using commercial microwave links: An overview of ongoing projects around the globe

Christian Chwala, Hidde Leijnse, Aart Overeem, Remko Uijlenhoet, Hagit Messer, Pinhas Alpert, Ali Doumounia, Martin Fencl, Vojtěch Bareš, Muhammad Sohail Afzal, Syed Hamid Hussain Shah, Giacomo

Roversi, Pier Paolo Alberoni, Remco van de Beek, Carlo De Michele, Michele D’Amico, Roberto Nebuloni and Harald Kunstmann

Karlsruhe Institute of Technology, KNMI, KNMI, Wageningen University & Research, Tel Aviv University, Tel Aviv University, Institut Des Sciences Université Ouagadougou, Czech Technical University in Prague, Czech Technical University in Prague, University of Agriculture Faisalabad, University of Agriculture Faisalabad,

MEEO s.r.l. and University of Bologna, ARPAE Emilia Romagna, SMHI, Politecnico di Milano (DICA), Politecnico di Milano (DEIB), Consiglio Nazionale delle Ricerche (EIIT) and Karlsruhe Institute of Technology

Abstract

Within the last ten years, researches from several countries have managed to get access to attenuation data from commercial microwave link (CML) networks to derive rainfall information. We given an overview of the currently ongoing CML research in Germany, the Netherlands, Israel, Burkina Faso, Czech Republic, Pakistan, Italy, and Sweden. We show selected results from each country and elaborate on the specific CML networks that are used. Furthermore we highlight the future perspectives of this techniques and discuss the crucial hurdles that have to be overcome.

Exploring the characteristics of mid-latitude precipitation: 3+ years of Micro Rain Radar observations from Stornoway, Scotland

Chris Kidd1 & Edward Graham2

1 Earth System Science Interdisciplinary Center, University of Maryland, College Park, and NASA/Goddard Space Flight Center, Greenbelt, USA. 2 Lews Castle College, University of the Highlands and Islands,

Stornoway, Scotland.

Abstract

The observation of precipitation over an extended period is a key element in understanding the characteristics of precipitation. In particular, the knowledge of the intensities and occurrences of precipitation are important for cross-checking the retrievals made by satellite sensors. In June 2015 a Micro Rain Radar was installed in Stornoway, NW Scotland (58.213°N, 6.398°W), with the aim of providing a long-term record of precipitation characteristics at a mid-/high-latitude site, one that is exposed to the weather systems of the North Atlantic. One of the main advantages of the MRR is the sensitivity to light precipitation, often experienced in this region, and the ability to contextualise the precipitation through analysis and interpretation of the vertical distribution of the precipitation, such as depth of precipitating system and freezing height. In addition, the nominal 1 minute temporal sampling (averaged from 10 second measurements) provides essential information on the occurrence of precipitation. This poster outlines and analyses the results from this study during these three years, providing an insight into the precipitation characteristics of this region, and how these characteristics compare with those of current satellite observations and retrievals.

Downscaled Atmospheric Forcings for Hyper-Resolution Land Surface Modeling

Tasnuva Rouf, Yiwen Mei, Viviana Maggioni, Paul Houser

George Mason University

Abstract

In this work we introduce physically-based downscaling techniques for surface meteorology to study surface flux, storage, and water balance changes and investigate the causality of these changes at the regional to local scale. A proof-of-concept has been implemented over the Oklahoma domain, where high-resolution ground-based observations are available for validation purposes. Hourly NLDAS (North America Land Data Assimilation System) forcing data (i.e., near-surface air temperature, pressure, humidity, wind speed, incident longwave and shortwave radiation, and precipitation) have been downscaled to 500m resolution over Oklahoma. The same techniques have been then applied at 1km resolution to the NASA MERRA-2 (Modern-Era Retrospective analysis for Research and Applications, Version 2) dataset in the more complex High Mountain Asia region, where ground validation is still a challenge. The precipitation downscaling approach is based on a nonparametric statistical scheme that uses a tree-based classification and a regression algorithm, in which atmospheric conditions, topography, and land cover are all considered as prognostic variables. Results are encouraging, showing that correlation coefficients and Nash-Sutcliffe indices between the downscaled dataset and ground observations are consistently higher (and biases are smaller) than the ones between the native resolution forcing data and the reference. The downscaled precipitation fields also show good spatial agreement with respect to several satellite precipitation products.

A daily 1°x1° gridded precipitation observational products database (ground-based, satellite and reanalysis datasets) in support of precipitation assessments

Rémy Roca1, Lisa Alexander2, Michael Bosilovich3, Stefan Contractor2, Adrien Guérou1, Rômulo A. J. Oliveira4, Gerald Potter3 and Sophie Cloché5

1) Laboratoire d'Etudes en Géophysique et Océanographie Spatiales (LEGOS, CNRS), Toulouse, France; 2) University of New South Wales, Sydney, Australia 3) NASA Goddard Space Flight Center, Greenbelt, Maryland

4) Geoscience Environnement Toulouse (GET, CNRS), Toulouse, France; 5) IPSL, Palaiseau, France

Abstract

Under the framework of the IPWG/GDAP Precipitation assessment and the WCRP Grand Challenge on Extremes, a multi-source database has been assembled and consistently formatted to facilitate the access and exploitation of a large ensemble of existing datasets. A common grid of 1x1 daily is used facilitate quantitative comparison among the disparate data products. In this poster, we will present the content of the database (currently composed of two land-only quasi global ground based data products, nine quasi global satellite products, three regional satellite products and five global reanalysis products). We will illustrate the preliminary investigation of the spread of this ensemble for various metrics. The emphasis will nevertheless be put on daily-based metrics including an assessment of the whole distribution but with special attention paid to the extremes. Early recommendations on how to use such an ensemble will be provided. Finally, an outlook for the evolution of the database will be offered for discussion with the IPWG community.

WRad: a long-term radiometric field campaign for characterizing w-band attenuation of precipitating clouds

D. Cimini(1,2), F. S. Marzano(2,3), L. Luini(4), C. Riva(4), M. Biscarini(2,3), L. Milani(2,3), K. De Sanctis(5), S. Di Fabio(2), F. Di Paola(1), E. Ripepi(1), E. Ricciardelli(1), F. Romano(1)

(1) CNR-IMAA, Italy; (2) CETEMPS, University of L’Aquila, Italy; (3) DIET, Sapienza University of Rome, Italy; (4) Politecnico di Milano, Milano, Italy; (5) HIMET, L’Aquila, Italy

Abstract

The effects of precipitating clouds on microwave propagation through the atmosphere have important applications in remote sensing and telecommunication. In particular, telecommunication applications (e.g., satellite communications) require an accurate estimation of the atmospheric attenuation to minimize the link outage probability. This is particularly important at W band (75–110 GHz), appearing to be the natural evolution of satellite communication systems, which is characterized by higher atmospheric losses (due to rain, liquid water clouds and atmospheric gases) with respect to lower frequencies. Thus, planning of future W band satellite communication systems requires experimental data to characterize all-sky atmospheric attenuation for the W-band radio channel. In this context, we propose WRad, a two-year field campaign exploiting ground-based microwave radiometers (MWR), satellite beacon receivers, weather radars, terrestrial radio links, disdrometers, raing gauges, and other meteorological in-situ instruments, for a complete characterization of W-band atmospheric attenuation in all-sky conditions, thus including precipitating clouds. In particular, WRad will exploit Sun-Tracking (ST) ground-based radiometry observations, which have been recently demonstrated for deriving atmospheric attenuation in all weather conditions [Marzano et al., 2016; Mattioli et al., 2017]. The WRad field campaign is proposed at two sites: the Milano “supersite” and the Potenza site. The Milano supersite includes the main campus of the Politecnico di Milano (PoliMI) and the experimental facility in Spino d’Adda, located ~25 km apart. The PoliMI main campus represents the core facility. Here is available a MWR with ST capability with four receiving channels at bands K (23.8 GHz), Ka (31.4 GHz), V (72.5 GHz) and W (82.5 GHz). In addition, the following relevant instruments are available: W- and D-band terrestrial links (73 and 83 GHz, 148 and 156 GHz, respectively), Ka- and Q-band beacon receivers (Alphasat satellite), GPS receiver, rain gauge, disdrometer, meteorological station. At the Spino d’Adda facility, a Ka- and V-band MWR with cosmic-back-ground (CB) capability is available, in addition to Alphasat beacon receivers, rain gauge, meteorological station. The Potenza site is located at the premises of CNR-IMAA in Tito Scalo near Potenza (~760 km from Milano). Here, a Ka- and V-band MWR with CB capability is available, in addition to GPS receiver, ceilometer, and meteorological station. The relative distance between the observations sites will be used to investigate site diversity and small/large scale spatial correlation of atmospheric attenuation.

The dataset collected during WRad will be used to characterize the attenuation of precipitating clouds at W band, generating distribution and diurnal/monthly statistics, which may serve as reference to radio-regulatory bodies. References: Marzano F. S., V. Mattioli, L. Milani, K. M. Magde and G. A. Brost, "Sun-Tracking Microwave Radiometry: All-Weather Estimation of Atmospheric Path Attenuation at Ka-, V-, and W-Band," in IEEE Transactions on Antennas and Propagation, vol. 64, no. 11, pp. 4815-4827, Nov. 2016. doi: 10.1109/TAP.2016.2606568 Mattioli V., L. Milani, K. M. Magde, G. A. Brost and F. S. Marzano, "Retrieval of Sun Brightness Temperature and Precipitating Cloud Extinction Using Ground-Based Sun-Tracking Microwave Radiometry," in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 10, no. 7, pp. 3134-3147, July 2017. doi: 10.1109/JSTARS.2016.2633439

Snowfall observations from the Space-borne perspective at Sub-millimeter frequencies

Xinxin Xie

Shanghai Spaceflight Institute of TT&C and Telecommunication

Abstract

Snowfall plays an important role in the climatology and meteorology due to their correlation to the global hydrology cycle and energy balance. However, lower density of snow particles and their complicated micro-physics, compared to raindrops, result in microwave channels lower than 100 GHz sometimes not optimal for snowfall detection, especially when talking about the ice-snowfall processes. To offer a deep insight into the micro-physical processes and micro-physical parameters of snowfall, sub-millimeter frequencies with polarized channels are proposed to fulfill this task. A detailed sensitivity study of brightness temperature to snow is carried out at the microwave frequencies ranging from 150 GHz to 900 GHz with the polarized microwave radiative transfer model. Considering that the snow particles are usually non-spherical and oriented, polarized signals caused by ice crystals with different shapes will be also investigated.

Improvement of COMS Rainfall Intensity using Passive Microwave Data

Yuha Kim, Jun Dong PARK, Geun-Hyeok RYU, and Jae-Dong JANG

Korea Meteorological Administration (KMA), Republic of Korea

Abstract

The core satellite of Global Precipitation Measurement (GPM) project had successfully launched on February 28 2014, and it is released with the newly produced GPM data together with the international constellation of research and operational satellites. In order to improve rainfall intensity (RI) of Communication, Ocean and Meteorological Satellite (COMS), more GPM GPROF precipitation products are introduced including DMSP-16/17/18, GCOM-W1, GPM core, MetOp-A/B, NOAA-18/19, and Suomi-NPP. The method is using simple relationship between precipitation from Passive Microwave (PMW) and COMS IR brightness temperature (TB). The probability matching method is applied to produce Look-up table between COMS IR TB and GPROF rain rates. COMS RI are varied with GPROF product, for example, when GMI precipitation is used, COMS RI seems strong than other sensors, and COMS IR from sounding sensors such as MetOP, NOAA shows relative low rain rate than imager sensors. Finally, it seems that new COMS IR has better correlation with ground measurement over the Korean peninsula. This COMS RI will be used in operational product. The comparisons between COMS RI and ground radar over Korea will be presented.

All-sky assimilation of IASI water vapor channels at KIAPS

Hyoung-Wook Chun, Jeon-Ho Kang, and Sihye Lee

KIAPS

Abstract

This presentation describes a trial of all-sky assimilation of IASI upper-tropospheric water vapor channels. The key parameters for all-sky assimilation are radiative calculation, observation error, bias correction, and quality control. All-sky brightness temperature (TB) and clear TB are calculated by RTTOV version 11.3 with the inputs of atmospheric profiles and hydrometers which are interpolated to observation space from KIAPS assimilation and forecast systems. The observation error is inflated with respect to cloud signals, i.e. the difference of all-sky TB and clear sky TB or the number of cloud contaminated channels. Bias correction has been performed with the off line method for clear sky condition. Our data assimilation system has 4 iterations of outer-loop. After finishing each step of the first, second, and third outer-loop, the analysis TBs are checked between the observed TB and background TB to consider the buddy check; e.g. the nearby measurements (buddies) tend to confirm each other. We are currently struggling to find the effect of all-sky assimilation of IASI water vapor channels in KIAPS cycling test.

All-sky assimilation of MHS in the Korea Integrated Model (KIM) system

Sihye Lee, Hyoung-Wook Chun, and Jeon-Ho Kang

Korea Institute of Atmospheric Prediction Systems (KIAPS), Seoul, Korea

Abstract

All-sky microwave radiance assimilation system has been developed in the Korea Integrated Model (KIM) analysis system. Initially, RTTOV-SCATT (version 11.3; Bauer et al., 2006; Geer et al., 2014) modules were implemented to assimilate the microwave humidity sounder (MHS) 183 GHz channels over ocean. A system development methodology for assimilating cloud and precipitation-affected satellite radiances has been developed in accordance with the ECMWF approach (Geer et al., 2014). There is no cloud or precipitation control variable, but in the minimization, cloud and precipitation are diagnosed from the dynamical and humidity fields every time-step including the first. In this study, the clear-sky and all-sky contributions to the assimilation of MHS are separated. The clear-sky assimilation shows the significant forecast benefits, but the all-sky assimilation is testing. Cloud and precipitation parameters are not directly assimilated in our system, but only temperature and humidity profiles are improved in the all-sky assimilation.

The Historical Database for Gridded Daily Precipitation Dataset over Latin America: The LatAmPrec Dataset

Daniel Vila, Dan Osgood, Sofía Martinez Saenz, and Manuel Bahn

CPTEC/INPE, and IRI

Abstract

A new dataset was commissioned over Latin America with the goal of supporting decision making in various socio-economic activities and in particular for climate insurance products. The Historical Database for Gridded Daily Precipitation Dataset over Latin America based on the Combined Scheme approach (hereafter, CoSch) developed at CPTEC/INPE, hereon referred to as LatAmPrec provides a new high-resolution, low-latency, gauge–satellite-based analysis of daily precipitation over Latin America for the period March 2000 – July 2017. In order to understand the strengths and limitations of the new dataset, two different validation methodologies were applied to assess the accuracy of the new product. The first one is based in a cross correlation process where independent gauges (randomly selected) are compared with the results of the CoSch algorithm using the remaining stations in the merging process. The second one, for climate insurance purposes, uses a statistical approach to assess how well the new dataset predicts evidence of bad years (years with climate impacts) across case studies. The evidence of bad years is based on farmer/expert focus groups, national datasets of climatic events and alternate remote sensing technologies. The results from both validation methodologies show that LatAmPrec has a good performance when compared with other data sources and it can capture satisfactorily the insurance-relevant losses on the ground. One of the main advantages of the new product is the high spatial resolution and the low latency when compared with other existing products used in the climate insurance industry.

Development of an active sensor module for the RTTOV-SCATT radiative transfer simulator

Fabrice Duruisseau, Philippe Chambon, Alan Geer

CNRM UMR 3589, Météo-France & CNRS, ECMWF

Abstract

Active microwave sensors are becoming widely used observations within the Numerical Weather Prediction community, either for validating model forecasts or for assimilation purposes. Like for the forward simulation of passive microwave observations, radar data simulations require to make assumptions on the scattering properties of hydrometeors. With the objective of simulating both active and passive microwave instruments within a single framework using the same radiative transfer assumptions into a widely-used tool in the NWP community, an active sensor module is currently under development within the RTTOV-SCATT software. The first simulations of the GPM/DPR instrument as well as the Cloudsat/CPR instrument with this simulator will be shown, based on the AROME model running operationally at Météo-France over five domains in the Tropics. In particular, some model biases highlighted with these first comparisons will be discussed.

Test Results for Version 6 IMERG

George J. Huffman 1; David T. Bolvin 1,2; Dan Braithwaite 3; Kuolin Hsu 3; Robert Joyce 4,5; Christopher Kidd 1,6; Eric J. Nelkin 1,2; Soroosh Sorooshian 3; Jackson Tan 1,7; Pingping Xie 5

1 NASA/GSFC, Greenbelt, MD, USA; 2 Science Systems and Applications, Inc., Lanham, MD, USA; 3 Univ. of California Irvine, Irvine, CA, USA; 4 Innovim, Greenbelt, MD, USA; 5 NOAA/NWS Climate Prediction Center, College Park, MD, USA; 6 Univ. of Maryland / ESSIC, College Park, MD, USA; 7 Universities Space Research

Assoc., Columbia, MD, USA

Abstract

The US Global Precipitation Measurement (GPM) mission science team merges intercalibrated observations from the international constellation of precipitation-relevant satellites to provide precipitation estimates on a 0.1° × 0.1° half-hourly grid in the Integrated Multi-satellitE Retrievals for GPM (IMERG) product. The next version, IMERG V06, is expected to be late in the test phase by meeting time and nearly ready for computation. As such, we plan to show early results that compare performance by V05 and V06 for the merged precipitation fields. In addition, we will show examples of the various additional data fields that IMERG provides, including the relatively new Quality Index. One key change from V05 to V06 is that the source of morphing vectors has been changed from the CPC 4-km global merged IR to the MERRA-2/GEOS-5 analyses/forecasts of vertically integrated vapor. This change permits morphing in the polar regions for the first time in IMERG, and we plan to show tests of how representative these data are. [But, we continue to mask out regions with snow and ice at the surface due to the limitations of the input passive microwave precipitation estimates, so these results only appear over open water and during summer land conditions.] We will also discuss the expected schedule for computing the entire record starting in 1998 with the Tropical Rainfall Measuring Mission (TRMM), including the eventual retirement of the legacy TRMM Multi-satellite Precipitation Analysis.

A Bayesian approach for merging multiple satellite soil moisture-based rainfall products with IMERG early run

(a) Christian Massari, (b)Viviana Maggioni, (a) Silvia Barbetta, (a)Luca Brocca, (c) Thierry Pellarin, (a) Paolo Filippucci, (a) Luca Ciabatta, (d) Yann H. Kerr, (e) Diego Fernández-Prieto

(a) National Research Council CNR, Research Institute for Geo-Hydrological Protection, Perugia, Italy; (b) George Mason University, Fairfax, VA, United States; (c) Univ. Grenoble Alpes, CNRS, IRD, Grenoble INP, IGE,

Grenoble F-38000, France; (d) Centre d’Etudes Spatiales de la BIOsphère (CESBIO), Université Toulouse 3 CNES CNRS IRD, Toulouse, France (e) European Space Agency (ESA), Frascati, Italy

Abstract

As a natural feature of the Earth’s weather system, rainfall is the main driver of the hydrological cycle. Rainfall plays an essential role in many applications including climate monitoring, extreme weather prediction and weather forecasting. On a global scale, ground-monitoring networks do not provide sufficient coverage and satellite rainfall products are often the only source of rainfall that guarantee a continuous temporal coverage. However, the indirect and the instantaneous nature of the measurement makes satellite rainfall products prone to errors (Kucera et al., 2013). Thanks to the strong connection between soil moisture and precipitation, capable to track accumulated precipitation estimates (rather than instantaneous), soil moisture can be successfully used to enhance the quality of satellite rainfall observations (Crow et al., 2011; Pellarin et al., 2013; Brocca et al. 2014). The SMOS+rainfall project of the European Space Agency (ESA), started in 2015 and concluded in 2017, has demonstrated the capability of the SMOS soil moisture product to enhance satellite rainfall information over land and has raised many interesting research questions related to the potential improvement that can be obtained by a combination of different soil moisture sensors. Here, we propose the use of a Bayesian approach for merging the Integrated Multi-Satellite Retrievals for GPM (IMERG early run) with multiple satellite rainfall products obtained from the inversion of the soil moisture retrievals derived from: 1) the Soil Moisture Active and Passive (SMAP) mission, 2) the Advanced Scatterometer (ASCAT) and 3) the Soil Moisture and Ocean Salinity (SMOS) mission via SM2RAIN (Brocca et al. 2014). Results in India, United States, Australia and Europe show that the simultaneous use of multiple soil moisture products through a Bayesian approach is able to increase the quality of IMERG early run product thus making it more appealing for hydrological applications.

1. Brocca, L., Ciabatta, L., Massari, C., Moramarco, T., Hahn, S., Hasenauer, S., Kidd, R., Dorigo, W., Wagner, W., Levizzani, V., 2014. Soil as a natural rain gauge: estimating global rainfall from satellite soil moisture data. J. Geophy. Res. 119 (9), 5128–5141. https://doi.org/10.1002/2014JD021489.

2. Crow, W.T., van den Berg, M.J., Huffman, G.J., Pellarin, T., 2011. Correcting rainfall using satellite-based surface soil moisture retrievals: the soil moisture analysis rainfall tool (SMART). Water Resour. Res. 47, W08521. 3. Kucera, P.A. , Ebert, E.E. , Turk, F.J. , Levizzani, V. , Kirschbaum, D. , Tapiador, F.J. , Loew, A. , Borsche, M. , 2013. Precipitation from space: advancing earth system science. Bull. Am. Meteorol. Soc. 94, 365–375. 4. Pellarin, T., Louvet, S., Gruhier, C., Quantin, G., Legout, C., 2013. A simple and effective method for correcting soil moisture and precipitation estimates using AMSR-E measurements. Remote Sens. Environ. 136, 28–36. https://doi.org/10. 1016/j.rse.2013.04.011.

SLALOM: An all-surface snow water path retrieval algorithm for the GPM Microwave Imager

Jean-François Rysman, Giulia Panegrossi, Paolo Sanò, Anna Cinzia Marra, Stefano Dietrich, Lisa Milani, and Mark S. Kulie

Institute of Atmospheric Sciences and Climate-CNR, NASA Goddard Space Flight Center, Michigan Technological University

Abstract

Snowfall plays a central role in the Earth radiative budget and water cycle. Therefore, there is a strong need to better characterise snowfall distribution and variability at the global scale. Passive microwave sensors, thanks to the sensitivity of their high-frequency channels to snowfall, could provide such a global information on snowfall. However, snowfall quantification using spaceborne microwave radiometers remains a challenging task due to the complex microphysics of snow clouds, and to the many sources of signal contamination (e.g., frozen background, presence of supercooled droplets). In this context, we developed an algorithm, named SLALOM, able to detect snowfall and to retrieve the associated snow water path (SWP) over any surface type using the GPM Microwave Imager (GMI). The algorithm is tuned and evaluated against coincident observations of the Cloud Profiling Radar (CPR) onboard CloudSat. It is composed of three modules for i) snowfall detection, ii) supercooled droplets detection and iii) SWP retrieval. This algorithm takes into account environmental conditions to retrieve SWP and does not rely on any surface classification scheme. The snowfall detection module is able to detect 83 % of snowfall events including light SWP (down to 1x10-3 kg.m-2) with a false alarm ratio of 0.12. The supercooled detection module detects 97 % of events with a false alarm ratio of 0.05. The SWP estimates shows a relative bias of -11 %, a correlation of 0.84 and a root mean square error of 0.04 kg.m -2. The SLALOM algorithm will be presented at IPWG 2018. In addition, several applications of SLALOM will be highlighted such as case studies of snowfall events and a 2-year high resolution 60°S-60°N snowfall occurrence distribution.

How well do triple-frequency radar signatures of simulated and observed melting snowflakes compare?

Davide Ori, Stefan Kneifel, Jose Dias Neto, and Mario Mech

University of Cologne, University of Cologne, University of Cologne, and University of Cologne

Abstract

The melting of ice particles produce distinctive radiative features at microwave frequencies such as the radar bright-band and increased signal attenuation. An accurate characterization of the scattering properties of melting ice particles is not only relevant for precipitation retrievals from space but also for utilizing the observational fingerprints, e.g. for model evaluation. While the number of scattering datasets for complex snow aggregates and rimed particles is rapidly increasing during recent years, the number of available scattering computations for realistic melting particles is very limited (especially regarding the number of particle sizes, shapes, melted fractions and frequencies included). This is certainly connected to the high complexity of the melting process and the large computational cost of scattering simulations. In this contribution, we use two scattering datasets of melting snowflakes to calculate triple-frequency radar signatures and linear depolarization ratios. We compare the range of simulated signatures with recent ground-based triple-frequency (X, Ka, W-Band) radar simulations for selected mid-latitude rainfall cases. The scattering simulations reveal an increased radar reflectivity and attenuation. The comparison between the two datasets shows some differences that are connected to the particle modeling algorithm. Although some features in the multifrequency space are well modeled by the scattering databases, none of them seems to be able to entirely match the range of observed values. Experiments with the scattering models provide some initial hints for the reason of the observed discrepancies between the models and the observations. The comparison with radar observations reveals that current scattering datasets of mixed-phase particles cannot yet fully represent the signature of melting snowflakes and additional work is needed to fully understand the relation between microphysical and scattering properties of mixed-phase precipitation.

The TAPEER 2.0 algorithm: a new implementation using the fully GPM Level-2 constellation (imagers and sounders)

Romulo A. Juca Oliveira, Marielle Gosset, Remy Roca, Adrien Guerou, Chris Kidd

GET/CNRS, IRD/GET/CNRS, LEGOS/CNRS, LEGOS/CNRS, ESSIC/University of Maryland/NASA

Abstract

The variety of passive microwave (PMW) radiometers (imagers and sounders) that compose the Global Precipitation Measurement (GPM) constellation enables a great opportunity to exploit and understand spatially and temporally the precipitation characteristics and its associated uncertainties. The Tropical Amount of Precipitation with an Estimate of ERrors (TAPEER) is now evolving to take full advantage of the GPM constellation of PMW imagers and sounders, and benefit from the latest evolution in the Level 2 products. The most recent algorithm implementation is based on the use of a larger amount of instantaneous rain rates retrievals based on the Goddard Profiling algorithm (GPROF) and the Precipitation Retrieval and Profiling Scheme (PRPS) algorithms. GPROF has been enhanced for GPM constellation and its actual version (GPROF2017) is applicable to different PMW radiometers, such as imagers (e.g., SSMI/S, GMI and AMSR-2) and sounders as well (i.e., MHS, NPP). In order to keep the full benefit of the Megha-Tropiques (MT) orbit over the tropics, GPROF PMW retrieval (which does not yet include SAPHIR) is complemented by PRPS retrieval on SAPHIR. As previously for TAPEER 1.5 algorithm, TAPEER 2.0 combines an estimation of the mean conditional rain rate and of the rainy fraction. TAPEER 2.0 allows to choose the constellation for calculating the daily rain fraction and to compute the conditional rain rate, independently. With this flexible approach, the use of sounders and imagers can thus be optimized according to the retrieval skills on each platform type. Sensitivity tests are currently carried out in order to finalize the choice of parameters in TAPEER 2.0. Our latest work on: evolving the error model for TAPEER and downloading the rainfall product by combining mechanistic and stochastic approaches, and the expected benefit for hydrological applications will also be discussed.

Status and outlook of the TAPEER 1.5 product

Sophie Cloché1, Rémy Roca2 and Marielle Gosset3

1) IPSL, Palaiseau, France 2) Laboratoire d'Etudes en Géophysique et Océanographie Spatiales (LEGOS, CNRS), Toulouse, France; 3) Geoscience Environnement Toulouse (GET, CNRS), Toulouse, France;

Abstract

TAPEER 1.5 is the current operational version of the constellation based daily precipitation estimates product of the Megha-Tropiques science team. It merges the instantaneous observations from various microwave imagers and the SAPHIR detection together with the fleet of geostationary IR imagery at full resolution. Since late 2011, it provides daily 1x1° estimates of precipitation together with an estimation of uncertainty. After recalling recent results showing the importance of the MT sampling and the good multi-year performances of the products, the status of the constellation and geo fleet is reviewed with special emphasis on new geostationary observations. We will present a confidence index has also been developed to help user assess the quality of the retrieval when missing data are identified. Data availability over the last 6 years will be reviewed together with a preliminary FAQ list oriented for the user. Finally, plan to handle the upcoming GOES-17 data and the possible need to use INSAT-3E data over the Indian Ocean will be discussed.

Rain clouds detection using the FengYun -3 Microwave Humidity Sounder II

Yang Guo

National Satellite Meteorological Center (NSMC)/CMA

Abstract

The microwave humidity sounder (MWHS) is a five channel microwave radiometer in the range of 150-191GHz onboard FY-3A and FY-3B. FY-3A and FY-3B were successfully launched in 2008 and 2010 respectively. The next generation of MWHS is a microwave humidity and temperature sounder (MWHS II). This sensor is a cross-track scanning instrument which observing in 15 channels at frequencies ranging from 89 to 191 GHz. Eight temperature sounding channels have center frequency at 118.75GHz oxygen absorption line, five humidity sounding channels have center frequency at 183.31GHz water vapor absorption line and two window channels centered at 89 and 150GHz. An important application of MWHS II data is atmospheric temperature and humidity profiles retrieval. These data also will be assimilated in the Global and Regional Assimilation and PrEdiction System (GRAPES) of China Meteorological Administration. Some MWHS II channels are sensitive to precipitation which are not used since the GRAPES is incapable of correctly assimilation these radiances. To determine the precipitation contamination for MWHS II radiance, we make use of a scattering index algorithm developed by Bennartz (1999). This method has only been applied over the sea to discrimination between precipitating and non-precipitating clouds. The statistical indexes, Probability Of Detection (POD) and Proportion Correct (PC), has been introduced to evaluate the rain clouds detection capability, discriminating between the raining and non-raining pixels between MWHS II and FlagPrecip from DPR data. Preliminary test suggests that the values of POD is 94% for MWHS II rain clouds detection.

The GIRAFE algorithm: early results and way forward

Rémy Roca2 , Sophie Cloché2, Adrien Guérou1, and Marc Schröder3

1) Laboratoire d'Etudes en Géophysique et Océanographie Spatiales (LEGOS, CNRS), Toulouse, France; 2) IPSL, Palaiseau, France 3) DWD, Offenbach, Germany

Abstract

In the framework on the Climate Monitoring SAF of EUMETSAT, a global climate-oriented product is under development. It will rely upon the Globally Interpolated Rain Fall Estimation (GIRAFE) algorithm and the aim is to provide global daily accumulation precipitation at 1°x1° resolution from 2002 on-wards, along with an estimation of uncertainty. The GIRAFE algorithm closely resembles the TAPEER one and merges the microwave constellation (sounders and imagers) with the geostationary infrared imagery up to 55° polewards and uses only the microwave data from 5° to the poles. A first integration has been performed using the GPROFv5 rainrates as an input and will be discussed at the conference. In particular, the sensitivity of the results to the role of the sounders will be highlighted. First attempts to estimate polar accumulation will be also shared. This early implementation is encouraging and paves the way towards more developments and the anticipated uncertainty modelling and validation procedure efforts will also be presented in the poster.

Classification of cloud top phase during winter monsoon over the Sea of Japan using A-train data

Munehisa K. Yamamoto1, Yu Asami2, and Shoichi Shige1

1: Graduate School of Science, Kyoto University, 2: Faculty of Science, Kyoto University

Abstract

A snowfall estimation scheme was implemented from the latest version (algorithm version 7) of GSMaP for GMI and SSMIS. This method enables to classify precipitation phase and estimate snowfall intensity. However, snowfall amount for GSMaP underestimates against ground radar data. A-train consists of multiple satellites carrying various instruments such as cloud radar, IR and microwave radiometers, and lidar. A recent study using A-train data pointed out that dominant microwave signature of snowfall is brightness temperature decrease caused by ice scattering rather than increase by cloud liquid water. This study correlates the horizontal distribution of cloud with the particle phase near the cloud top to improve snowfall estimation combining microwave and infrared radiometers. Over the Sea of Japan, northwestern winter monsoon winds receive heat and moisture from ocean, and cumulus snowfall clouds frequently develop. We compared horizontal cloud pattern from MODIS with cloud particle phase from CALIOP in the typical snow cloud case. In the initial stage of the development, the phase near the cloud top is liquid cloud water. As cloud development progresses, the phase changes from liquid to solid. Horizontal pattern of cloud changes from shallow-uniform to sparse distribution. We classified the phases using area-average and homogeneity from MODIS data.

GEO-KOMPSAT-2A Precipitation Retrieval Algorithm: Status and Ongoing Works

Dong-Bin Shin and Damwon So

Yonsei University

Abstract

A precipitation retrieval algorithm has been developed for the Advanced Meteorological Imager (AMI) onboard the GEO-KOMPSAT-2A (GK-2A), the second Korea’s geostationary satellite. The algorithm uses the a-priori information including the microwave rainfall data from the low-earth orbiting satellites and infrared (IR) brightness temperatures from geostationary satellites. The algorithm can better perform with a variety of a-priori information describing all possible precipitating systems. In addition, separation of physically different precipitating systems likely to improve the accuracy of retrieval process. However, it has been well known that such the separation can be hardly achieved based on the measurements of cloud top temperatures. This algorithm tries to utilize the radiative characteristics observed differently for different wavelengths in IR spectral regions. The characteristics include the different emissivity as a function of wavelength and cloud thickness. Using the brightness temperature differences (BTDs) between IR channels the algorithm discriminates five types of precipitating clouds: one shallow and four non-shallow types. In addition to the separation of cloud types in the databases, the algorithm also uses databases classified by latitudinal bands. The bands are separated with four latitudinal zones. The separation of database based on latitudes may have an effect of distinguishing the cloud types that can occur regionally. The a-priori databases are thus classified with 20 different categories. Once the a-priori databases are constructed, the algorithm inverts the AMI IR brightness temperatures to the surface rainfall rate based on a Bayesian approach. The Bayesian approach has advantages on using multi-channel brightness temperatures simultaneously and utilizing the probability of rainfall reserved in the a-priori databases. As a proxy for the AMI this algorithm first tests the AHI data. The retrieval results and the status and ongoing works of the algorithm development will be discussed.

Development of a quantitative precipitation nowcasting algorithm for AMI onboard the GEO-KOMPSAT-2A satellite

Yu-Ri Lee, Dong-Bin Shin, and Sukbum Hong

Yonsei University, Kyung Hee University

Abstract

Statistical approachs for quantitative precipitation nowcasts (QPNs) have emerged with recent advances in sensors in geostationary orbits, which provide more frequent observations at higher spatial resolutions. Several investigations have reported that extrapolation-based prediction can be more effective than numerical weather model-based prediction for a short lead time of less than a few hours. This study introduces an algorithm based on extrapolation method to produce QPN products for Advanced Meteorological Imager (AMI) onboard South Korea’s second geostationary satellite (GEO-KOMPSAT-2A or GK-2A). The algorithm first provides the 3-hour potential accumulated rainfall (PAR) for a very-short-range forecast using extrapolation and then calculates the probability of rainfall (POR) during the same time period through statistical methods. The algorithm for PAR first identifies rainfall features and computes the motion vectors of the identified rainfall features through tracking between two consecutive GK-2A rain rate images. The algorithm can then extrapolate future rainfall rates from current and previous rainfall rates outputs. In addition, the extrapolated precipitation fields are used as inputs for the POR estimation process, which produces the rainfall probability during the same 3-h period. Retrieved PAR and POR from the algorithm are validated with 3-h accumulated rainfalls obtained from the GK-2A rain rate products. Forecasting accuracy of the PAR algorithm with time has also been investigated. The accuracy tends to decrease with increasing lead time, as expected. It may be due to the limitations of the extrapolation technique and degradation of the accuracy of the PAR estimates. The validation results will be discussed.

Estimation of cloud top heights from two geostationary satellites

Jonghyuk Lee and Dong-Bin Shin

Yonsei University

Abstract

We have developed a stereoscopy-based algorithm to estimate Cloud Top Height (CTH) using visible images from two geostationary satellites. This algorithm is based on the parallax, which is a displacement of the apparent position of a specific cloud target. The algorithm starts with remapping two visible images (their acquisitions are nearly simultaneous) to a common grid so that two images are comparable. Then, the parallax of the cloud from its real position can be identified. CTH is finally estimated by trigonometric calculation using the parallax and the geometry information by which two satellite sensors observe the same cloud target (i.e. satellite zenith angle). In this study, visible images from FY-2E and Himawari-8 were used; in near future, FY-2E visible image will be changed to that of FY-4A, and GK-2A visible image will also applied to this algorithm. This presentation provides details of algorithm development.

The Rain No-Rain Classification near Coastline over East Asia

Hwayoung JEOUNG1, Geun-Hyeok RYU1, Eun-kyoung Seo2, Jun Dong PARK1, and Jae-Dong JANG1

1Korea Meteorological Administration (KMA), Republic of Korea, 2Kongju National University, Republic of Korea

Abstract

The passive microwave rainfall algorithm based on the Bayesian inversion is built with the optimized a-prior database in order to reflect specifics of rainfall systems occurring in East Asia (Seo et al, 2016). It has been built separately between land and ocean around East Asia because different physical principles are used over each surface. In this study, rain no-rain classification over coast is focused and introduced. Over coast, the microwave footprint is a mixture of ocean and land surface. So the rain no-rain classification (RNC) is applied in two steps. In the first step, land RNC based on Grody (1991) is optimized with the satellite TBs of GMI (GPM) with respect to radar rainfall estimates in 2014 over the Korean peninsula. And then the pixels pass through the second RNC which use the relationship between TB23V and PCT89 (Spencer, 1989). For finding the best cutoff in the second RNC, the spatial distribution of PCT89 (GMI) and rainfall (Radar) was checked for optimization 𝛽𝛽 of PCT89. Using statistical analysis, Heidke skill score (HSS), the rain cutoff threshold (𝑇𝑇𝐵𝐵23−𝑃𝑃𝐶𝐶𝑇𝑇89(𝛽𝛽=0.55)>-35) is derived. For the validation based on Radar, the result shows that the probability of detection (POD) is 0.58 and false alarm ratio (FAR) is 0.18 and Heidke skill score (HSS) is 0.56 from July, 2016 to November, 2017.

Temporal evolution of vertical structure of precipitation

Ziad S. Haddad

Jet Propulsion Laboratory, California Institute of Technology

Abstract

The satellite measurements that are relied upon to make estimates of surface precipitation are actually much more sensitive to the condensed water in the upper portion of the cloud than to the water that is sedimenting out of the bottom of the cloud. Yet all of today's global precipitation products attempt to estimate surface rain only, with no explicit estimates of the condensed-water amounts or vertical distribution above the surface. This gap is particularly egregious when one considers the fairly dense temporal sampling of today's radiometer constellation, which allows the tracking of storms in time, if the spatial characteristics were estimated. Recent analyses have laid the foundation for the production of such vertical structure information from the radiometer constellation, demonstrating the ability to estimate cloud-top heights (or rather "condensed-water Heights") for different condensed-mass thresholds, as well as the first two vertical principal components vPC1 and vPC2 of condensed mass, and the total Condensed Water Path. While these quantities are of course mutually correlated, it takes at least 3 PCs of brightness temperature to capture 99% of the variability for any given radiometer in the constellation, so there are at least three independent pieces of information in the radiances from any one radiometer beam, and these different pieces of information are useful for different interpretations of the condensed water within the column sensed by the beam. The main reason that all this information is important, however relatively coarse its level of vertical detail, is that it provides systematic estimates of the spatial structure of the cloud and, when tracked in time, its evolution. Such cloud-structure information is crucial to analyze storm processes. The only current alternative for such contextual data are the geostationary-IR cloud-tracking algorithms -- however, these suffer from two shortcomings: they do not agree among one another in the segmentation or the tracking (with coarser segmentation producing inconsistent lifecycle statistics though it is a more consistent indicator of storm organization and intensification); and they are not sensitive to the water below the very top of the cloud. In this poster, different options to produce 4-dimensional storm data from satellite microwave radiometers are compared and the results illustrated from actual observations.

A precipitation retrieval algorithm over oceans using low-resolution microwave emission signatures

Ji-Hye Kim and Dong-Bin Shin

Sejong University and Yonsei University

Abstract

Passive microwave emission signatures from rain water in precipitating clouds are more physically related to surface rainfall than scattering signals from ice particles. The emission signatures are, however, available at relatively low resolutions with the currently operating microwave radiometers’ specifications. This study develops a precipitation retrieval algorithm using only microwave emission signatures to produce level-3 data over the global ocean for climate applications. The algorithm constructs a priori databases (DBs) on a lower spatial resolution (0.5°×0.5° latitude-longitude grids), which shows better linearity in the brightness temperature (TB)-rain rate (R) relation than higher resolutions. The atmospheric profiles and surface conditions in the a priori DBs are based on the National Centers for Environmental Prediction (NCEP) Climate Forecast System Reanalysis (CFSR) data. The NCEP CFSR data include all variables needed for simulating microwave observations, except for the rain water profile. The rain water profiles are thus simulated separately by adjusting the rain water profiles estimated by the Global Precipitation Measurement (GPM) satellite to match the corresponding emission signals. The algorithm has been tested for the Special Sensor Microwave Imager/Sounder (SSMIS) and simulation of TBs has been performed by the Radiative Transfer for the Television and Infrared Observation Satellite (TIROS) Operational Vertical Sounder (RTTOV) model. Moreover, the a priori DBs are classified into 11 types of sub-DBs based on the Köppen climate classification criteria to better include regional precipitation characteristics. The inversion module of the algorithm is based on a Bayesian approach. The selection of a proper DB in the retrieval is first processed by a Support Vector Machines (SVM) method to discriminate the classified climate regimes. The freezing level and near sea surface temperature included in the reanalysis tag are then additionally used to limit the retrieval range in the selected sub-DB. Comparisons of the retrieval results with other level-3 products will be discussed.

Hybrid methodology for precipitation estimation using Hydro-Estimator over Brazil

Ricardo Siqueira, Daniel Vila

CPTEC/INPE

Abstract

Rainfall measurement is a very important topic to society and for the understanding of the weather and climate therefore, needs to be calculated as accurately as possible. Counteracting the problem of the high temporal and spatial variability of precipitation, geostationary satellites sensors have been proved an excellent tool to this task, providing scans with fast temporal resolution and detecting the growth and decay of rain cells. Using infra-red (IR) images obtained from the Geostationary Operational Environmental Satellites (GOES), the Hydro-Estimator (HYDRO) algorithm produces instantaneous precipitation estimates with 30 minutes temporal resolution and 4 km spatial resolution with a very low latency compared with other more sophisticated methodologies (i.e., passive microwave-based algorithms). However, the IR algorithm has some limitations to estimate precipitation on some cloud systems. In order to to overcome this problem, the main objective of this study is to develop a light and fast algorithm, based on the histogram matching (HM) technique, to combine the superior sampling and low latency of the HYDRO IR product with more accurate active microwave-based products using data obtained over Brazil. The combined adjusted HYDRO (AHYDRO) product was validated against Brazil rain gauge network for two years (2016-2017) and the performance was assessed by using standard statistical metrics and categorical index. Results show that the HM technique is able to minimize the large variability and discrepancies among HYDRO and observed precipitation over Brazil. At same time, is able to generate a better bias performance while maintaining the same correlation levels before the adjustment.

What is the value of radiometer measurements in conjunction with multi-frequency radar data?

S. Joseph Munchak, Ian S. Adams, and Mircea Grecu

NASA Goddard Space Flight Center, Morgan State University

Abstract

Radar and passive microwave radiometer are the pre-eminent tools used to remotely sense precipitation from space and aircraft, for mapping applications and process investigations. Owing to their separate heritage, differing viewing geometries and resolutions, and often the unavailability of both types of measurements simultaneously from the same platform, algorithms for the separate instruments have reached greater maturity than those that combine the passive and active measurements. In recent years, advancement in combined active-passive algorithms has been spurred by the Global Precipitation Measurement (GPM) mission, which derives precipitation profiles from Ku/Ka-band radar and multi-frequency radiometer measurements that are used in Bayesian retrievals across a constellation of passive sensors, necessitating physical agreement between the active and passive measurements. This work will draw from the GPM combined algorithm as well as algorithms designed for use with airborne datasets to quantify the additional information provided by passive microwave radiometer measurements to multi-frequency radar in different regimes (e.g., land/ocean, convective/stratiform). The impact on retrieved integral and profile quantities as well as surface parameters will be presented. One particular quantity that benefits from the combination of slant-viewing radiometer and near-nadir radar measurements is a parameter that describes the bulk aspect ratio in the ice phase. With the next generation of cloud and precipitation observing satellites in the pre-formulation phase, suggestions for future instrument passive and active frequency combinations and viewing geometries will be given.

Passive and Active Microwave Transfer (PAMTRA)

Mario Mech, Maximilian Maahn, Stefan Kneifel, Davide Ori, Susanne Crewell, and Pavlos Kollias

Universität zu Köln, Cologne, Germany, Cooperative Institute for Research in Environmental Sciences, University of Colorado Boulder, Boulder, CO, USA, NOAA Earth System Research Laboratory, Boulder, CO,

USA, School of Marine and Atmospheric Sciences, Stony Brook University, NY, USA

Abstract

Forward models are a key tool to compare observations and models by converting the output of atmospheric numerical models to synthetic observations. They are also an integral part of inversion algorithms that aim to retrieve geophysical variables of interest from observations. Here, the comprehensive microwave forward model PAMTRA (Passive and Active Microwave TRAnsfer) is introduced, which can simulate passive and active measurements across the microwave spectral region up to 800 GHz. The passive forward model in PAMTRA provides up- and down-welling polarized brightness temperatures and radiances for arbitrary observation angles, while the active forward simulator is capable of simulating radar Doppler spectra and their moments. Both can be applied to arbitrary plane-parallel atmospheric scenes, including those with complex hydrometeor assumptions. PAMTRA implements various gas absorption models and methods for the approximation of the scattering properties (Mie, T-matrix, DDA, self-similar Rayleigh-Gans) and uses the same for the passsive and active forward simulations. The core module is written in FORTRAN90, whereas the framework and user interface are python based. Therefore, it the model is easy to use and extendable. Furthermore, various applications from recent studies where PAMTRA has been applied will be shown.

The HOAPS Climatology V4: updates and results from comparisons to various satellite-based products with focus on precipitation

a) Marc Schröder, a) Karsten Fennig, a) Thomas Spangehl, a) Marloes Gutenstein, b) Jörg Burdanowitz, b) Christian Klepp, c) Stephan Bakan, a) Axel Andersson, a) Kathrin Graw

a) Deutscher Wetterdienst, b) University of Hamburg, c) Max-Planck-Institute for Meteorology

Abstract

The global water cycle is a key component of the global climate system as it describes and links many important processes such as evaporation, convection, cloud formation and precipitation. Through latent heat release, it is also closely connected to the global energy cycle and its changes. The difference between precipitation and evaporation yields the freshwater flux, which indicates if a particular region of the Earth receives more water through precipitation than it loses through evaporation or vice versa. On global scale and long time periods, however, the amounts of evaporation and precipitation are balanced. A profound understanding of the water cycle is therefore a key prerequisite for successful climate modelling. The Hamburg Ocean Atmosphere Parameters and Fluxes from Satellite Data (HOAPS) set is a fully satellite based climatology of precipitation, evaporation, freshwater budget and other variables over the global ice free oceans. All geophysical parameters are derived from passive microwave radiometers, except for the SST, which is taken from AVHRR measurements. Starting with the release 3.1, the HOAPS climate data record is hosted by the EUMETSAT Satellite Application Facility on Climate Monitoring (CM SAF) and the further development is shared with the University of Hamburg and the MPI-M. The latest version of the HOAPS data set, version 4, includes the SSM/I and SSMIS records and uncertainty estimates for parameters related to evaporation. These HOAPS data products are available as monthly averages and 6-hourly composites on a regular latitude/longitude grid with a spatial resolution of 0.5° x 0.5° from July 1987 to December 2014. Covering nearly 28 years the new HOAPS data set is highly valuable for climate applications. The data can be retrieved from the CM SAF web user interface http://wui.cmsaf.eu and from http://www.hoaps.org. The presentation will cover an overview of HOAPS 4, recent enhancements and an application example. A 1D-Var retrieval scheme was implemented and adapted successfully for, among others, total column water vapour and is currently further adapted to estimate liquid water path and precipitation. We show results from comparisons to various other data records with a specific focus on the assessment of the stability and precipitation. Sound uncertainty estimates are provided for flux-related parameters and first results are available for precipitation as well. Finally, we show results from an analysis of freshwater flux and water vapour transport which will be jointly analysed with sea surface salinity in near future.

Precipitation trends in East Africa from an ensemble of IR-based satellite products

Elsa Cattani (1), Vincenzo Levizzani (1), and Andrés Merino (2)

(1) CNR-ISAC, Bologna, Italy, (2) Environmental Institute, University of León, Spain

Abstract

The Africa Rainfall Climatology (ARC) v2, Climate Hazards Group InfraRed Precipitation with Stations (CHIRPS) v2, and Tropical Applications of Meteorology using SATellite (TAMSAT) African Rainfall Climatology And Time Series (TARCAT) v3 satellite rainfall products are exploited to study the spatial and temporal variability of East Africa (EA) rainfall between 1983 and 2017 through the time series of selected rainfall indices from the joint CCl/CLIVAR/JCOMM Expert Team on Climate Change Detection and Indices (ETCCDI). The indexes total rainfall amount (PRCPTOT), Simple Daily Intensity (SDII), number of precipitating days (R1), number of consecutive dry and wet days (CDD and CWD), and number of very heavy precipitating days (R20) were analyzed. The scope of the work is to draw the attention on the rainfall trend and variability identifying significant trend patterns regardless of the single satellite product, and also estimating the trend rate variability stemming from the multiplicity of the satellite products. The trend spatial patterns are recognized through the Mann-Kendall technique, considering the time series of the ensemble mean of the three satellite products and the corresponding time series of the standard deviations, which were interpreted as error bars associated with the ensemble mean time series. Indications on rainfall trends were extracted at annual and seasonal scales and the regions that more frequently exhibit statistically significant trends are located in eastern Kenya, Somalia at the border with western Ethiopia, northern Tanzania, and limited areas of South Sudan. At the seasonal scale increasing trends were identified for the October-November-December PRCPTOT, SDII, and R20 indices over eastern EA, with the exception of central Kenya, where rising trends with limited areas of significance stand out for R1 and CWD, distinguishable also at the yearly scale. In March-April-May rainfall decline is perceivable only through R1 and CWD in particular over the eastern EA region, whereas PRCPTOT, even though associated with negative trends, does not present any high confidence areas.

Characteristics of heavy ice precipitation in extratropical cyclones observed by GPM/DPR

Shizuka Akiyama(1), Shoichi Shige(1), Munehisa K. Yamamoto(1), Toshio Iguchi(2), Michael P. Bauer(3, 4)

1)Kyoto University, 2)NICT, 3)Columbia University, 4)NASA Goddard Institute for Space Studies

Abstract

Heavy Ice Precipitation flags (HIPs), indicating the existence of large ice particles such as hails, graupels and snowflakes, have been introduced by Iguchi et al. (2018) into GPM/DPR products based on Dual Frequency Ratio (DFR). The purpose of this study is to reveal the feature of the horizontal and vertical distribution and the ice particle density of HIP using DPR. Because high frequency of HIP is seen in the storm track in winter, we focus on extratropical cyclones for three oceanic regions (North Pacific Ocean, North Atlantic Ocean and south hemisphere) respectively. In all regions, HIPs occur frequently poleward of the cyclone center far within 500km. HIPs are more often seen with cyclones over the north Atlantic Ocean than others. We also investigated in detail a case of oceanic extratropical cyclone in March 2015. Here we found scattered HIP in the cold side of cold front and line-shape HIP of tens kilometers wide along the bent-back front direction. The probability density functions in KuZm-DFRm plane and contoured frequency by altitude diagrams (CFADs) for the cyclone suggests that the particle density of line-shape HIP decreases from the storm top toward the surface. This is probably due to aggregation of snowflakes in the lower layer. On the other hand, scatter-shape HIP has the value of high DFRm and KuZm near the storm top. This may be caused by updrafts raising a large size of snowflakes or the attenuation of Ka-band by supercooled water above the storm top.

Are the satellite precipitation estimation algorithms performing well in extreme rainfall situations in Brazil?

Rayana Palharini and Daniel Vila

INPE

Abstract

Extreme rainfall is one of the most severe weather hazards to affect all the globe. Some works use statistical approaches for interpreting extreme events, to choose a constant threshold based on the empirical distribution of the variable at each location is a way that it ensures that a given fraction of events will by definition be extreme. The most common way to choose a threshold is to use the technique of quantiles that are cut points dividing the range of a probability distribution into contiguous intervals with equal probabilities but there is still no consensus in meteorology, about the precipitation thresholds for the identification of extreme events, and is therefore quite variable. The present work intends to show on the role of extreme rainfall events impact on micro-climates of the selected regions in Brazil. In this analysis, the assessment of the frequency and intensity of these events occurring at these regions is of fundamental importance. It is expected to answer if the current techniques of satellite precipitation estimates are capable of accurately identifying extreme events occurred in Brazil and to quantify the uncertainties of the retrievals. The importance of this work lies in the fact that investigating different rainfall retrieval techniques and their limitations in estimating extreme rainfall events can show which techniques need to be improved and to estimates the biggest sources of uncertainties. Such investigations are important to boost the development of new techniques and promote improvements in the already existing ones. These improvements in the techniques are of great importance to help risk managers to get a quick response and save lives during extreme events. Satellite precipitation estimates for extreme events still a challenging problem and the identification associated with these events in Brazil is relevant.

Analysis of heavy rainfall events occurred in Italy by using Microwave and Infrared Technique

E. Ricciardelli(1), F. Di Paola(1), S. Gentile(1,2), A. Cersosimo(1), D. Cimini(1,2), D. Gallucci(1), E. Geraldi(1), S. Larosa(1), S. Nilo(1), E. Ripepi(1), F. Romano(1), M. Viggiano(1)

(1) Institute of Methodologies for Environmental Analysis, National Research Council (CNR-IMAA), 85050 Tito Scalo (PZ), Italy; (2) CETEMPS, University of L’Aquila, 67100 L’Aquila, Italy

Abstract

On 9 and 10 September 2017 violent rainstorm events affected the centre of Italy. The brunt of the flooding took the city of Livorno where more than 200 mm of precipitation was recorded in few hours causing the overflow of Rio Maggiore that led tragic events and did great harms. The case study is analysed using Weather Research and Forecast (WRF) model [1] and two algorithms based on satellite observations: the Precipitation Evolving Technique (PET) [2,4] and the Rain Class Evaluation from Infrared and Visible observations (RainCEIV) [3] algorithm. The RainCEIV rain classes and PET rain rate results have been compared with the rain rates estimated by the Italian Weather Radar Network. The statistical assessment reveals a good performance for both the algorithms (for PET: bias=1.06, POD=0.85, FAR=0.20; for RainCEIV: bias=1.28, POD=0.83, FAR=0.35). WRF is able to forecast the event, though with errors in actual structure, location, and time. For this reason, the combined use of different observational tools could support the WRF simulation to provide a better characterization of the event. 1. Skamarock, W. C., and Coauthors A Description of the Advanced Research WRF Version 3. NCAR Technical Note NCAR/TN-475+STR, 2008, doi:10.5065/D68S4MVH 2. Di Paola, F., Casella, D., Dietrich, S., Mugnai, A., Ricciardelli, E., Romano, F., and Sanò, P.: Combined MW-IR Precipitation Evolving Technique (PET) of convective rain fields, Nat. Hazards Earth Syst. Sci., 12, 3557-3570, doi:10.5194/nhess-12-3557-2012, 2012. 3. Ricciardelli, E., Cimini, D., Di Paola, F., Romano, F., and Viggiano, M.: A statistical approach for rain intensity differentiation using Meteosat Second Generation–Spinning Enhanced Visible and InfraRed Imager observations, Hydrol. Earth Syst. Sci., 18, 2559-2576, doi:10.5194/hess-18-2559-2014, 2014. 4. Di Paola, F., Ricciardelli, E., Cimini, D., F., Romano, F., Viggiano, M., and Cuomo, V.: Analysis of Catania Flash Flood Case Study by Using Combined Microwave and Infrared Technique, J. Hydrometeorology, 1989-1998,2014.

Hydrological performances of satellite-based rainfall products over Europe

Stefania Camici(1), Luca Ciabatta(1), Christian Massari(1), Viviana Maggioni(2), Luca Brocca(1)

(1)Research Institute for Geo-Hydrological Protection, National Research Council, Perugia, Italy (s.camici@irpi.cnr.it) (2)Department of Civil, Environmental, and Infrastructure Engineering, George Mason

University (GMU), Fairfax, VA 22030, USA

Abstract

Rainfall is the primary input for hydrologic models that simulate the rainfall-runoff processes at basin scale. Because rainfall is highly variable in space and time, accurate hydrological simulations require accurate rainfall data at the best possible resolution. The conventional rain gauge observations in many parts of the world are sparse and unevenly distributed. An alternative to traditional rain gauge observations could be satellite-based rainfall products (SRPs) that nowadays are available on a global scale at ever increasing spatial and temporal resolution. This study proposes a comprehensive assessment of SRPs for flood modeling in Europe. For this purpose, multiple SRPs (i.e., the Tropical Rainfall Measurement Mission (TRMM) Multi-satellite Precipitation Analysis TMPA; the Climate Prediction Center (CPC) Morphing algorithm, CMORPH, the Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks, PERSIANN) will be used to force different lumped hydrologic models (e.g., MISDc, HBV, HYMOD) over multiple (+900) basins throughout Europe with different sizes and physiographic characteristics. As outcome of this study, guidelines for the optimal use of SRPs in flood modeling will be drawn. In particular, this study will allow to: 1) assess the quality of different SRPs for flood modelling and its relationship with climatic/geomorphological conditions; 2) explore the connection between the accuracy of SRPs and their performance in terms of flood modeling taking into account the rainfall-runoff model structure as well.