scholarly journals Projection Pursuit Regression (PPR) on Statistical Downscaling Modeling for Daily Rainfall Forecasting

2021 ◽  
Vol 5 (2) ◽  
pp. 326-332
Author(s):  
Rio Pradani Putra ◽  
Dian Anggraeni ◽  
Alfian Futuhul Hadi
Keyword(s):  
Observational Data ◽  
Statistical Downscaling ◽  
General Circulation ◽  
Regional Scale ◽  
Projection Pursuit ◽  
Predictor Variable ◽  
Circulation Model ◽  
Rainfall Forecasting ◽  
Projection Pursuit Regression ◽  
Response Variable

Rainfall forecasting has an important role in people's lives. Rainfall forecasting in Indonesia has complex problems because it is located in a tropical climate. Rainfall prediction in Indonesia is difficult due to the complex topography and interactions between the oceans, land and atmosphere. With these conditions, an accurate rainfall forecasting model on a local scale is needed, of course taking into account the information about the global atmospheric circulation obtained from the General Circulation Model (GCM) output. GCM may still be used to provide local or regional scale information by adding Statistical Downscaling (SD) techniques. SD is a regression-based model in determining the functional relationship between the response variable and the predictor variable. Rainfall observations obtained from the Meteorology Climatology and Geophysics Council (BMKG) are a response variable in this study. The predictor variable used in this study is the global climate output from GCM. This research was conducted in a place, namely Kupang City, East Nusa Tenggara because it has low rainfall. The Projection Pursuit Regression (PPR) will be used in this SD method for this study. In PPR modeling, optimization needs to be done and model validation is carried out with the smallest Root Mean Square Error (RMSE) criteria. The expected results must have a pattern between the results of forecasts and observations showing or approaching the observational data. The PPR model is a good model for predicting rainfall because The results of the forecast and observation show that the results of the rainfall forecast are observational data.

PREDICTORS AND THEIR DOMAIN FOR STATISTICAL DOWNSCALING OF CLIMATE IN BANGLADESH

2016 ◽  
Vol 78 (6-12) ◽  
Author(s):  
Mahiuddin Alamgir ◽  
Sahar Hadi Pour ◽  
Morteza Mohsenipour ◽  
M. Mehedi Hasan ◽  
Tarmizi Ismail
Keyword(s):  
Statistical Downscaling ◽  
General Circulation ◽  
Large Scale ◽  
Regional Scale ◽  
Circulation Model ◽  
Local Scale ◽  
Mitigation Measures ◽  
Grid Points ◽  
Hydrological Services ◽  
Atmospheric Variables

Reliable projection of future rainfall in Bangladesh is very important for the assessment of possible impacts of climate change and implementation of necessary adaptation and mitigation measures. Statistical downscaling methods are widely used for downscaling coarse resolution general circulation model (GCM) output at local scale. Selection of predictors and their spatial domain is very important to facilitate downscaling future climate projected by GCMs. The present paper reports the finding of the study conducted to identify the GCM predictors and demarcate their climatic domain for statistical downscaling in Bangladesh at local or regional scale. Twenty-six large scale atmospheric variables which are widely simulated GCM predictors from 45 grid points around the country were analysed using various statistical methods for this purpose. The study reveals that large-scale atmospheric variables at the grid points located in the central-west part of Bangladesh have the highest influence on rainfall.  It is expected that the finding of the study will help different meteorological and agricultural organizations of Bangladesh to project rainfall and temperature at local scale in order to provide various agricultural or hydrological services.

Pemanfaatan CFSRv2 untuk Statistical Downscaling menggunakan Principal Component Regression dan Partial Least Square

2019 ◽  
Vol 8 (1) ◽  
Author(s):  
Khairunnisa Khairunnisa ◽  
Rizka Pitri ◽  
Victor P Butar-Butar ◽  
Agus M Soleh
Keyword(s):  
Statistical Downscaling ◽  
General Circulation ◽  
Principal Component Regression ◽  
Circulation Model ◽  
Principal Component ◽  
Meteorological Station ◽  
Grid Size ◽  
Partial Least Square ◽  
Least Square ◽  
Output Data

This research used CFSRv2 data as output data general circulation model. CFSRv2 involves some variables data with high correlation, so in this research is using principal component regression (PCR) and partial least square (PLS) to solve the multicollinearity occurring in CFSRv2 data. This research aims to determine the best model between PCR and PLS to estimate rainfall at Bandung geophysical station, Bogor climatology station, Citeko meteorological station, and Jatiwangi meteorological station by comparing RMSEP value and correlation value. Size used was 3×3, 4×4, 5×5, 6×6, 7×7, 8×8, 9×9, and 11×11 that was located between (-40) N - (-90) S and 1050 E -1100 E with a grid size of 0.5×0.5 The PLS model was the best model used in stastistical downscaling in this research than PCR model because of the PLS model obtained the lower RMSEP value and the higher correlation value. The best domain and RMSEP value for Bandung geophysical station, Bogor climatology station, Citeko meteorological station, and Jatiwangi meteorological station is 9 × 9 with 100.06, 6 × 6 with 194.3, 8 × 8 with 117.6, and 6 × 6 with 108.2, respectively.

scholarly journals Temperature Projections over the Indus River Basin of Pakistan Using Statistical Downscaling

Atmosphere ◽  
2021 ◽  
Vol 12 (2) ◽  
pp. 195
Author(s):  
Muhammad Saleem Pomee ◽  
Elke Hertig
Keyword(s):  
21St Century ◽  
Statistical Downscaling ◽  
General Circulation ◽  
Winter Season ◽  
Circulation Model ◽  
Indus Basin ◽  
Indus River ◽  
Temperature Changes ◽  
Basin Scale ◽  
Maximum Heating

We assessed maximum (Tmax) and minimum (Tmin) temperatures over Pakistan’s Indus basin during the 21st century using statistical downscaling. A particular focus was given to spatiotemporal heterogeneity, reference and General Circulation Model (GCM) uncertainties, and statistical skills of regression models using an observational profile that could significantly be improved by recent high-altitude observatories. First, we characterized the basin into homogeneous climate regions using K-means clustering. Predictors from ERA-Interim reanalysis were then used to model observed temperatures skillfully and quantify reference and GCM uncertainties. Thermodynamical (dynamical) variables mainly governed reference (GCM) uncertainties. The GCM predictors under RCP4.5 and RCP8.5 scenarios were used as “new” predictors in statistical models to project ensemble temperature changes. Our analysis projected non-uniform warming but could not validate elevation-dependent warming (EDW) at the basin scale. We obtained more significant warming during the westerly-dominated seasons, with maximum heating during the winter season through Tmin changes. The most striking feature is a low-warming monsoon (with the possibility of no change to slight cooling) over the Upper Indus Basin (UIB). Therefore, the likelihood of continuing the anomalous UIB behavior during the primary melt season may not entirely be ruled out at the end of the 21st century under RCP8.5.

scholarly journals Palaeogeographic controls on climate and proxy interpretation

2016 ◽  
Vol 12 (5) ◽  
pp. 1181-1198 ◽  
Author(s):  
Daniel J. Lunt ◽  
Alex Farnsworth ◽  
Claire Loptson ◽  
Gavin L. Foster ◽  
Paul Markwick ◽  
...  
Keyword(s):  
Global Change ◽  
Carbon Cycle ◽  
Ocean Circulation ◽  
General Circulation ◽  
Regional Scale ◽  
Circulation Model ◽  
Adjustment Factor ◽  
Climate Signal ◽  
Proxy Records ◽  
Global Mean

Abstract. During the period from approximately 150 to 35 million years ago, the Cretaceous–Paleocene–Eocene (CPE), the Earth was in a “greenhouse” state with little or no ice at either pole. It was also a period of considerable global change, from the warmest periods of the mid-Cretaceous, to the threshold of icehouse conditions at the end of the Eocene. However, the relative contribution of palaeogeographic change, solar change, and carbon cycle change to these climatic variations is unknown. Here, making use of recent advances in computing power, and a set of unique palaeogeographic maps, we carry out an ensemble of 19 General Circulation Model simulations covering this period, one simulation per stratigraphic stage. By maintaining atmospheric CO2 concentration constant across the simulations, we are able to identify the contribution from palaeogeographic and solar forcing to global change across the CPE, and explore the underlying mechanisms. We find that global mean surface temperature is remarkably constant across the simulations, resulting from a cancellation of opposing trends from solar and palaeogeographic change. However, there are significant modelled variations on a regional scale. The stratigraphic stage–stage transitions which exhibit greatest climatic change are associated with transitions in the mode of ocean circulation, themselves often associated with changes in ocean gateways, and amplified by feedbacks related to emissivity and planetary albedo. We also find some control on global mean temperature from continental area and global mean orography. Our results have important implications for the interpretation of single-site palaeo proxy records. In particular, our results allow the non-CO2 (i.e. palaeogeographic and solar constant) components of proxy records to be removed, leaving a more global component associated with carbon cycle change. This “adjustment factor” is used to adjust sea surface temperatures, as the deep ocean is not fully equilibrated in the model. The adjustment factor is illustrated for seven key sites in the CPE, and applied to proxy data from Falkland Plateau, and we provide data so that similar adjustments can be made to any site and for any time period within the CPE. Ultimately, this will enable isolation of the CO2-forced climate signal to be extracted from multiple proxy records from around the globe, allowing an evaluation of the regional signals and extent of polar amplification in response to CO2 changes during the CPE. Finally, regions where the adjustment factor is constant throughout the CPE could indicate places where future proxies could be targeted in order to reconstruct the purest CO2-induced temperature change, where the complicating contributions of other processes are minimised. Therefore, combined with other considerations, this work could provide useful information for supporting targets for drilling localities and outcrop studies.

scholarly journals Utility of daily vs. monthly large-scale climate data: an intercomparison of two statistical downscaling methods

2007 ◽  
Vol 4 (5) ◽  
pp. 3413-3440 ◽  
Author(s):  
E. P. Maurer ◽  
H. G. Hidalgo
Keyword(s):  
Statistical Downscaling ◽  
General Circulation ◽  
Large Scale ◽  
Impact Analysis ◽  
Climate Model ◽  
Daily Precipitation ◽  
Circulation Model ◽  
Reanalysis Data ◽  
Temperature Extremes ◽  
Daily Data

Abstract. Downscaling of climate model data is essential to most impact analysis. We compare two methods of statistical downscaling to produce continuous, gridded time series of precipitation and surface air temperature at a 1/8-degree (approximately 140 km² per grid cell) resolution over the western U.S. We use NCEP/NCAR Reanalysis data from 1950–1999 as a surrogate General Circulation Model (GCM). The two methods included are constructed analogues (CA) and a bias correction and spatial downscaling (BCSD), both of which have been shown to be skillful in different settings, and BCSD has been used extensively in hydrologic impact analysis. Both methods use the coarse scale Reanalysis fields of precipitation and temperature as predictors of the corresponding fine scale fields. CA downscales daily large-scale data directly and BCSD downscales monthly data, with a random resampling technique to generate daily values. The methods produce comparable skill in producing downscaled, gridded fields of precipitation and temperatures at a monthly and seasonal level. For daily precipitation, both methods exhibit some skill in reproducing both observed wet and dry extremes and the difference between the methods is not significant, reflecting the general low skill in daily precipitation variability in the reanalysis data. For low temperature extremes, the CA method produces greater downscaling skill than BCSD for fall and winter seasons. For high temperature extremes, CA demonstrates higher skill than BCSD in summer. We find that the choice of most appropriate downscaling technique depends on the variables, seasons, and regions of interest, on the availability of daily data, and whether the day to day correspondence of weather from the GCM needs to be reproduced for some applications. The ability to produce skillful downscaled daily data depends primarily on the ability of the climate model to show daily skill.

scholarly journals The CSIRO Mk3L climate system model version 1.0 – Part 1: Description and evaluation

2011 ◽  
Vol 4 (2) ◽  
pp. 483-509 ◽  
Author(s):  
S. J. Phipps ◽  
L. D. Rotstayn ◽  
H. B. Gordon ◽  
J. L. Roberts ◽  
A. C. Hirst ◽  
...  
Keyword(s):  
Sea Ice ◽  
General Circulation Model ◽  
Land Surface ◽  
General Circulation ◽  
Regional Scale ◽  
Circulation Model ◽  
Climate System ◽  
System Model ◽  
Desktop Computer ◽  
Climate System Model

Abstract. The CSIRO Mk3L climate system model is a coupled general circulation model, designed primarily for millennial-scale climate simulations and palaeoclimate research. Mk3L includes components which describe the atmosphere, ocean, sea ice and land surface, and combines computational efficiency with a stable and realistic control climatology. This paper describes the model physics and software, analyses the control climatology, and evaluates the ability of the model to simulate the modern climate. Mk3L incorporates a spectral atmospheric general circulation model, a z-coordinate ocean general circulation model, a dynamic-thermodynamic sea ice model and a land surface scheme with static vegetation. The source code is highly portable, and has no dependence upon proprietary software. The model distribution is freely available to the research community. A 1000-yr climate simulation can be completed in around one-and-a-half months on a typical desktop computer, with greater throughput being possible on high-performance computing facilities. Mk3L produces realistic simulations of the larger-scale features of the modern climate, although with some biases on the regional scale. The model also produces reasonable representations of the leading modes of internal climate variability in both the tropics and extratropics. The control state of the model exhibits a high degree of stability, with only a weak cooling trend on millennial timescales. Ongoing development work aims to improve the model climatology and transform Mk3L into a comprehensive earth system model.

scholarly journals Role of Nonlinear Atmospheric Response to SST on the Asymmetric Transition Process of ENSO

2009 ◽  
Vol 22 (1) ◽  
pp. 177-192 ◽  
Author(s):  
Masamichi Ohba ◽  
Hiroaki Ueda
Keyword(s):  
Observational Data ◽  
Zonal Wind ◽  
General Circulation ◽  
Circulation Model ◽  
La Niña ◽  
Atmospheric Response ◽  
Cold Event ◽  
Transition Processes ◽  
La Nina ◽  
Sst Anomalies

Abstract Physical processes that are responsible for the asymmetric transition processes between El Niño and La Niña events are investigated by using observational data and physical models to examine the nonlinear atmospheric response to SST. The air–sea coupled system of ENSO is able to remain in a weak, cold event for up to 2 yr, while the system of a relatively warm event turns into a cold phase. Through analysis of the oceanic observational data, it is found that there is a strong difference in thermocline variations in relation to surface zonal wind anomalies in the equatorial Pacific (EP) during the mature-to-decaying phase of ENSO. The atmospheric response for the warm phase of ENSO causes a rapid reduction of the EP westerlies in boreal winter, which play a role in hastening the following ENSO transition through the generation of upwelling oceanic Kelvin waves. However, the anomalous EP easterlies in the cold phase persist to the subsequent spring, which tends to counteract the turnabout from the cold to warm phase of ENSO. A suite of idealized atmospheric general circulation model (AGCM) experiments are performed by imposing two different ENSO-related SST anomalies, which have equal amplitudes but opposite signs. The nonlinear climate response in the AGCM is found at the mature-to-decaying phase of ENSO that closely resembles the observations, including a zonal and meridional shift in the equatorial positions of the atmospheric wind. By using a simple ocean model, it is determined that the asymmetric responses of the equatorial zonal wind result in different recovery times of the thermocline in the eastern Pacific. Thus, the differences in transition processes between the warm and cold ENSO event are fundamentally due to the nonlinear atmospheric response to SST, which originates from the distribution of climatological SST and its seasonal changes. By including the asymmetric wind responses the intermediate air–sea coupled model herein demonstrates that the essential elements of the redevelopment of La Niña arise from the nonlinear atmospheric response to SST anomalies.

Statistical downscaling of general circulation model outputs to precipitation, evaporation and temperature using a key station approach

2016 ◽  
Vol 7 (4) ◽  
pp. 683-707
Author(s):  
D. A. Sachindra ◽  
F. Huang ◽  
A. Barton ◽  
B. J. C. Perera
Keyword(s):  
General Circulation Model ◽  
Statistical Downscaling ◽  
General Circulation ◽  
Emission Scenario ◽  
Circulation Model ◽  
Reanalysis Data ◽  
Maximum Temperature ◽  
Monthly Precipitation ◽  
Regression Equations ◽  
Highly Correlated

Using a key station approach, statistical downscaling of monthly general circulation model outputs to monthly precipitation, evaporation, minimum temperature and maximum temperature at 17 observation stations located in Victoria, Australia was performed. Using the observations of each predictand, over the period 1950–2010, correlations among all stations were computed. For each predictand, the station which showed the highest number of correlations above 0.80 with other stations was selected as the first key station. The stations that were highly correlated with that key station were considered as the member stations of the first cluster. By employing this same procedure on the remaining stations, the next key station was found. This procedure was performed until all stations were segregated into clusters. Thereafter, using the observations of each predictand, regression equations (inter-station regression relationships) were developed between the key stations and the member stations for each calendar month. The downscaling models at the key stations were developed using reanalysis data as inputs to them. The outputs of HadCM3 pertaining to A2 emission scenario were introduced to these downscaling models to produce projections of the predictands over the period 2000–2099. Then the outputs of these downscaling models were introduced to the inter-station regression relationships to produce projections of predictands at all member stations.

Sign in / Sign up

Export Citation Format

Share Document