Patrick Laux | Karlsruhe Institute of Technology (KIT) (original) (raw)

Papers by Patrick Laux

This study investigates the ability of the regional climate model Weather Research and Forecastin... more This study investigates the ability of the regional climate model Weather Research and Forecasting (WRF) in simulating the seasonal and interannual variability of hydro-meteorological variables in the Tana River basin (TRB) in Kenya, East Africa. The impact of two different land use classifications , i.e., the Moderate Resolution Imaging Spectroradiometer (MODIS) and the US Geological Survey (USGS) at two horizontal resolutions (50 and 25 km) is investigated. Simulated precipitation and temperature for the period 2011–2014 are compared with Tropical Rainfall Measuring Mission (TRMM), Climate Research Unit (CRU), and station data. The ability of Tropical Rainfall Measuring Mission (TRMM) and Climate Research Unit (CRU) data in reproducing in situ observation in the TRB is analyzed. All considered WRF simulations capture well the annual as well as the inter-annual and spatial distribution of precipitation in the TRB according to station data and the TRMM estimates. Our results demonstrate that the increase of horizontal resolution from 50 to 25 km, together with the use of the MODIS land use classification, significantly improves the precipitation results. In the case of temperature, spatial patterns and seasonal cycle are well reproduced, although there is a systematic cold bias with respect to both station and CRU data. Our results contribute to the identification of suitable and regionally adapted regional climate models (RCMs) for East Africa.

Especially in semi-arid or arid regions, where rainfall is limited to only few months per year, r... more Especially in semi-arid or arid regions, where rainfall is limited to only few months per year, rainfall is the most important factor affecting crop growth and yield. Every year, farmers are faced with the crucial question when to start planting. Do the first rainfalls after the dry season resemble the Onset of the Rainy Season (ORS) or not? The farmers' seeds and effort will be lost if no major rainfall follows within the following weeks. It is apparent, that the choice of the planting date is crucial for crop yield. A fuzzy logic algorithm for estimating the onset of the rainy season and the optimal planting date is developed. It is based on rainfall data and accounts for agriculturally meaningful aspects, expressed in terms of the following definition constraints: i) rainfall amounts, ii) number of rainy days, and iii) the occurrence of dry spells at the beginning of the growing season. The ORS algorithm, which is calculating the planting date for each year, is coupled to the physically based crop model CropSyst. A Monte Carlo approach is applied to generate annual planting dates (1979-2003). Therefore, the definition constraints, which are allowed to vary within reasonable parameter ranges, are generated randomly. The averaged crop yield is serving as performance measure for each realization. The parameter range of the best realizations is retained. Various iterations are necessary to obtain a robust set of definition parameters. The coupled ORS definition-crop modelling system is applied for different crop species and observation sites across Cameroon for the period 1979-2003. It is shown that the derived optimal planting dates would allow significantly increased crop yields compared to the existing planting rules. Finally, based on the robust definition parameterizations, expected future crop yield is estimated by statistical downscaling of different global climate scenarios (ECHAM5). Keywords: Crop Modelling, Monte Carlo, Onset of the Rainy Season, Planting date

High Performance Computing in Science and Engineering ´15, 2016

Egu General Assembly Conference Abstracts, Apr 9, 2013

Monthly and seasonal meteorological forecasts are routinely produced by several international wea... more Monthly and seasonal meteorological forecasts are routinely produced by several international weather services using global coupled ocean-atmosphere general circulation models. This kind of information can be used as source of information in operational hydrological monitoring and forecasting systems to improve early drought warnings. In March 2011, a new version of the global coupled model of the National Centre for Environmental Prediction, the Climate Forecast System (CFS) Version 2, became operational providing real-time ensemble forecasts up to nine months. However, a comprehensive analysis of the CFS forecast for the prediction of droughts in water stress regions has not yet been performed. In this study we evaluate the CFS precipitation forecasts for large-scale droughts that occurred during the rainy season in West Africa, East Africa and India. The target areas are large-scale river-basins like Volta (West Africa), Ganges (India) and the administrative area of Kenya. The forecasts are compared to monthly precipitation observations provided on a regular grid by the Global Precipitation Climatology Centre. In addition, the CFS performance is evaluated using areal monthly precipitation amount of the river basin of interest as an indicator for dry months. The verification is done for the period 1982-2009 using all ensemble members of the retrospective CFS archive. The outcomes of this study illustrate, that the CFS in some cases can simulate general features of the monthly precipitation regime for the respective river basins. However, an evaluation using the entire retrospective CFS forecasts demonstrates a low accuracy. Furthermore, the seasonal forecasts of monthly precipitation are characterized by a large over- and underestimation during the rainy season depending on the target region. In this presentation, the following issues are highlighted: (i) The performance of the CFS precipitation forecast for individual events such as the severe India drought in 2007 and the Sahel drought in 1983; (ii) The CFS forecast performance for predicting areal monthly precipitation of a river basin for different lead times using a set of verification measures to determine bias, accuracy and skill; (iii) The value of the CFS forecast if the monthly areal information is used for a warning of dry months during the rainy season.

Different circulation pattern (CP) classification methodologies are applied for the Jordan region... more Different circulation pattern (CP) classification methodologies are applied for the Jordan region. The obtained daily CP time series are statistically tested for mutual dependency using the χ2-test. The magnitude of the dependency is assessed by means of the adjusted contingency coefficient and Cramér's coefficient, separately for the whole-year-round and the seasonal consideration. The persistence of the mutual dependencies is analyzed using a moving-window approach for the period 1961-1990. In order to estimate the possibility of making predictions of a certain CP classification, the Guttman's λ is calculated. The different methodologies are tested for usability for CP conditional rainfall modelling. Most of the mutual CP classification approaches are found to be non-independent. The highest correlation occurs between Beck's and Alpert's classification approach for the whole-year-round consideration, different results are obtained for the seasonal consideration. The strength of the mutual dependencies between the different classifications is found to depend strongly on the season. The greatest dependencies exist for winter, the lowest for summer. The relationships are found to remain relatively stable over the analyzed period. The performance of the different classification schemes for rainfall modelling within the Jordan region is analyzed. Except for two observation stations in the southern part of the research area, all the CP conditional approaches are superior to the unconditional rainfall modelling. The semi-objective Alpert CP classification is found to perform slightly better than the fully objective methodologies. Keywords circulation pattern analysis; Jordan region; χ2-test; Cramer's V; adjusted contingency coefficient; Guttman's λ

ABSTRACT The development and application of statistical techniques with a special focus on a reca... more ABSTRACT The development and application of statistical techniques with a special focus on a recalibration of meteorological or hydrological forecasts for an elimination of systematic differences (bias) between forecasts and observations has received a great deal of attention in recent years. The objective of this study is to propose several statistical techniques with different degree of complexity and to evaluate and compare their performance for a recalibration of seasonal ensemble forecasts of monthly precipitation. We use retrospective forecasts of the second version of the Climate Forecast System (CFS2) which are compared to monthly observations provided by the Global Precipitation Climatology Centre (GPCC). The target region is the Volta basin in West Africa. The CFS forecast are characterized by strong biases in comparison to the GPCC observations. In this poster presentation first results of this investigation are presented using three straightforward recalibration techniques applied for the ensemble mean.

In this paper, the frequencies of droughty and wet weather patterns are investigated. For this re... more In this paper, the frequencies of droughty and wet weather patterns are investigated. For this reason, a multi objective fuzzy rule - based classification method has been applied. This classification conditions daily rainfall time series to large-scale atmospheric weather patterns. First, ...

Several recent works have shown a controversy about the reliability of non-urban weekly cycles ov... more Several recent works have shown a controversy about the reliability of non-urban weekly cycles over large-scales in Europe. For example, Sanchez-Lorenzo et al. (2008) found weekly cycles for several climatic variables in Spain, but this work was criticized by Hendricks Fransen et al. (2009). In Central Europe, using different climatic variables and several time periods, Bäumer and Vogel (2007) and Laux and Kunstmann (2008) showed significant annual weekly cycles over Germany. Contrarily, Hendricks Franssen (2008) and Barmet et al. (2009) did not find any significant annual weekly cycles over Switzerland. These two latter works mainly focused their analysis on precipitation, which is well-known as a climatic variable with high variability, and consequently it is more difficult to detect any significant change in their series. In this work we present a seasonal analysis of a dataset with long-term series available in Switzerland and Germany, covering the major part of the 20th century...

ABSTRACT Monthly and seasonal meteorological forecasts are routinely produced by several internat... more ABSTRACT Monthly and seasonal meteorological forecasts are routinely produced by several international weather services using global coupled ocean-atmosphere general circulation models. This kind of information can be used as source of information in operational hydrological monitoring and forecasting systems to improve early drought warnings. In March 2011, a new version of the global coupled model of the National Centre for Environmental Prediction, the Climate Forecast System (CFS) Version 2, became operational providing real-time ensemble forecasts up to nine months. However, a comprehensive analysis of the CFS forecast for the prediction of droughts in water stress regions has not yet been performed. In this study we evaluate the CFS precipitation forecasts for large-scale droughts that occurred during the rainy season in West Africa, East Africa and India. The target areas are large-scale river-basins like Volta (West Africa), Ganges (India) and the administrative area of Kenya. The forecasts are compared to monthly precipitation observations provided on a regular grid by the Global Precipitation Climatology Centre. In addition, the CFS performance is evaluated using areal monthly precipitation amount of the river basin of interest as an indicator for dry months. The verification is done for the period 1982-2009 using all ensemble members of the retrospective CFS archive. The outcomes of this study illustrate, that the CFS in some cases can simulate general features of the monthly precipitation regime for the respective river basins. However, an evaluation using the entire retrospective CFS forecasts demonstrates a low accuracy. Furthermore, the seasonal forecasts of monthly precipitation are characterized by a large over- and underestimation during the rainy season depending on the target region. In this presentation, the following issues are highlighted: (i) The performance of the CFS precipitation forecast for individual events such as the severe India drought in 2007 and the Sahel drought in 1983; (ii) The CFS forecast performance for predicting areal monthly precipitation of a river basin for different lead times using a set of verification measures to determine bias, accuracy and skill; (iii) The value of the CFS forecast if the monthly areal information is used for a warning of dry months during the rainy season.

 Several recent works have generated a controversy about the reliability of non-urban weekly cyc... more  Several recent works have generated a controversy about the reliability of non-urban weekly cycles over large-scales in Europe (see Section 3 in XY83 Poster for more details).

High Performance Computing in Science and Engineering ‘13, 2013

International Journal of Climatology, 2012

Several statistical downscaling techniques are intercompared and evaluated with respect to daily ... more Several statistical downscaling techniques are intercompared and evaluated with respect to daily station-based precipitation in the eastern Mediterranean/Middle East region. The study introduces unconditioned and precipitationconditioned SANDRA (Simulated ANnealing and Diversified RAndomization) cluster analysis (SCA) as new downscaling approaches and additionally uses the two widely used techniques of canonical correlation analysis (CCA) and multiple linear regression analysis (MR). For the precipitation-conditioned SANDRA cluster analysis different weights (percentages of contribution to the clustering) are evaluated. Furthermore, two different predictor combinations are used, a simple one only including mean sea level pressure (SLP), and a more complex one additionally including 500 hPa-geopotential heights, 500 hPa-vorticity and 1000 hPa-moisture flux. Analyses are carried out on a daily basis for the main rainy season from November to March for the period . It is shown that SLP, as single predictor, does not perform sufficiently well, but adding further predictors considerably improves model performance in terms of increased explained variance and model stability as well as reduced root mean square error (RMSE). From all selected techniques MR and CCA show the best performance for the SLP-based models, with comparable results for both techniques, whereas precipitationconditioned SANDRA cluster analysis performs best when further predictors are included. Performance differences between all techniques are generally smaller than those for a particular technique using different predictor sets.

... Affiliation: AA(Karlsruhe Institute of Technology (KIT), Institute for Meteorology and Climat... more ... Affiliation: AA(Karlsruhe Institute of Technology (KIT), Institute for Meteorology and Climate Research, D-82467 Garmisch-Partenkirchen, Germany, patrick.laux@kit.edu), AB(Karlsruhe Institute of Technology (KIT), Institute for Meteorology and Climate Research, D-82467 ...

In such regions, where rainfall is limited to only few months per year, the exact determination o... more In such regions, where rainfall is limited to only few months per year, the exact determination of the rainy seasons' onset is of crucial interest for farming management. Every year, farmers are faced with the question when to start sowing. Do the first rainfalls after the dry season ...

High Performance Computing in Science and Engineering ´15, 2016

Egu General Assembly Conference Abstracts, Apr 9, 2013

High Performance Computing in Science and Engineering ‘13, 2013

International Journal of Climatology, 2012