Found 2 publication(s)
Thies, B. (2008): A novel day/night-technique for area-wide precipitation retrieval over Central Europe using MSG SEVIRI data Philipps-University-Marburg, phd thesis
- view metadata
- DOI: 10.17192/z2008.0906
- Abstract: Knowledge of the spatio-...
- Keywords: | Cloud properties | rain retrieval | precipitation | rainfall | Optical sensors |
Abstract:Knowledge of the spatio-temporal precipitation distribution is of great value in agriculture, water engineering, climatology and risk management. So far, no adequate method existed for the detection and monitoring of precipitation at high temporal and spatial resolutions in most parts of the world where radar networks are not available. Due to spectral constraints, existing retrieval techniques rely on a relationship between rainfall probability and intensity and the cloud top temperature measured in an infrared channel. These techniques show considerable drawbacks concerning precipitation processes in the mid-latitudes. Improved techniques for rain area identification based on spectral enhancements of new generation satellite systems used to be only available on polar orbiting platforms with poor temporal resolutions. Furthermore, these algorithms are only applicable during day-time. With the advent of Meteosat Second Generation (MSG) Spinning-Enhanced Visible and InfraRed Imager (SEVIRI) in 2004, a geostationary satellite system with significantly improved spectral and spatial resolutions has become available. The central aim of the present study therefore was to develop a novel method for operational precipitation detection during day- and night-time based on MSG SEVIRI data. The focus of the newly developed scheme lies on precipitation processes in the mid-latitudes in connection with extra-tropical cyclones. It is therefore not only applicable to convectively dominated rain areas but also to precipitating cloud areas of advective-stratiform character. The newly developed rainfall retrieval scheme based on the advanced second-generation GEO system MSG SEVIRI rests upon the following conceptual model: • Precipitating cloud areas are characterized by a sufficiently high cloud water path and ice particles in the upper part. • Cloud areas with higher rainfall intensities are characterized by a higher cloud water path and a higher amount of ice particles in the upper part. • Convective clouds with very high rainfall intensities are characterized by a large vertical extension and a high rising cold cloud top. Based on this conceptual design, the new retrieval scheme consists of an entirely new methodology compiling novel and innovative algorithms and approaches. The following three components are the focal parts of the novel technique: • A new algorithm for the identification of the rain area during day- and night-time was developed for SEVIRI. The method allows not only a proper detection of mainly convective rain areas but also enables the detection of advective-stratiform precipitation (e.g. in connection with mid-latitude frontal systems). It is based on information about the CWP and the cloud phase in the upper cloud regions. • An infrared retrieval technique appropriate for convective precipitation processes in the mid-latitudes was successfully transferred and adapted to MSG SEVIRI. The phenomenon of positive brightness temperature differences between the WV and IR channels (dTWV-IR), which enables the detection and classification of convectively dominated raining cloud areas was investigated for the WV and IR channels of SEVIRI. Based on radiative transfer calculations, which revealed the existence of positive ΔTWV-IR for all SEVIRI WV-IR differences, the dTWV technique could be applied and transferred to SEVIRI. • A new technique for precipitation process and rainfall intensity separation was developed for SEVIRI. The process separation and the further subdivision relies on information about the cloud top height, the cloud water path and the cloud phase in the upper parts. The subdivision is realized in a stepwise manner. In a first step the rain area is separated into the subareas of convective and advective-stratiform precipitation processes. In the following both separated process areas are divided into subareas of differing rainfall intensities. The process separation and the subdivision of the convective precipitation area relies on information about the cloud top height. The subdivision of the advective-stratiform precipitation area is based on information about the CWP and the particle phase in the upper parts of the cloud. The rain area and the process-oriented rainfall intensities detected and classified by the newly developed retrieval technique were validated against corresponding ground-based radar data of Germany, representative for mid-latitude precipitation processes. The results of the validation study indicate persuading performance of the new algorithm concerning rain area identification as well as process and intensity differentiation and indicate the stability of the introduced conceptual design.
Kühnlein, M.; Appelhans, T.; Thies, B. & Nauss, T. (2014): Improving the accuracy of rainfall rates from optical satellite sensors with machine learning — A random forests-based approach applied to MSG SEVIRI. Remote Sensing of Environment 141, 129–143.
- view metadata
- DOI: 10.1016/j.rse.2013.10.026
- Abstract: The present study aims t...
- Keywords: | MSG-SEVIRI | rainfall retrieval | Rainfall rate | Random forests | Maschine learning | Geostationary satellites | Optical sensors |
Abstract:The present study aims to investigate the potential of the random forests ensemble classification and regression technique to improve rainfall rate assignment during day, night and twilight (resulting in 24-hour precipitation estimates) based on cloud physical properties retrieved from Meteosat Second Generation (MSG) Spinning Enhanced Visible and InfraRed Imager (SEVIRI) data. Random forests (RF) models contain a combination of characteristics that make them well suited for its application in precipitation remote sensing. One of the key advantages is the ability to capture non-linear association of patterns between predictors and response which becomes important when dealing with complex non-linear events like precipitation. Due to the deficiencies of existing optical rainfall retrievals, the focus of this study is on assigning rainfall rates to precipitating cloud areas in connection with extra-tropical cyclones in mid-latitudes including both convective and advective-stratiform precipitating cloud areas. Hence, the rainfall rates are assigned to rain areas previously identified and classified according to the precipitation formation processes. As predictor variables water vapor-IR differences and IR cloud top temperature are used to incorporate information on cloud top height. ?T8.7–10.8 and ?T10.8–12.1 are considered to supply information about the cloud phase. Furthermore, spectral SEVIRI channels (VIS0.6, VIS0.8, NIR1.6) and cloud properties (cloud effective radius, cloud optical thickness) are used to include information about the cloud water path during daytime, while suitable combinations of temperature differences (?T3.9–10.8, ?T3.9–7.3) are considered during night-time. The development of the rainfall rate retrieval technique is realised in three steps. First, an extensive tuning study is carried out to customise each of the RF models. The daytime, night-time and twilight precipitation events have to be treated separately due to differing information content about the cloud properties between the different times of day. Secondly, the RF models are trained using the optimum values for the number of trees and number of randomly chosen predictor variables found in the tuning study. Finally, the final RF models are used to predict rainfall rates using an independent validation data set and the results are validated against co-located rainfall rates observed by a ground radar network. To train and validate the model, the radar-based RADOLAN RW product from the German Weather Service (DWD) is used which provides area-wide gauge-adjusted hourly precipitation information. Regarding the overall performance, as indicated by the coefficient of determination (Rsq), hourly rainfall rates show already a good correlation with Rsq = 0.5 (day and night) and Rsq = 0.48 (twilight) between the satellite and radar based observations. Higher temporal aggregation leads to better agreement. Rsq rises to 0.78 (day), 0.77 (night) and 0.75 (twilight) for 8-h interval. By comparing day, night and twilight performance it becomes evident that daytime precipitation is generally predicted best by the model. Twilight and night-time predictions are generally less accurate but only by a small margin. This may due to the smaller number of predictor variables during twilight and night-time conditions as well as less favourable radiative transfer conditions to obtain the cloud parameters during these periods. However, the results show that with the newly developed method it is possible to assign rainfall rates with good accuracy even on an hourly basis. Furthermore, the rainfall rates can be assigned during day, night and twilight conditions which enables the estimation of rainfall rates 24 h day.