Kuala Lumpur, Malaysia, 8th-11th November 2012 Climate Extremes © Crown copyright Met Office

Contents What is ‘Extreme’ and why use indices? Calculating Extremes using CDO Cautionary Note: Comparing Extremes in Model and Observed data This presentation is the combination of three different parts: the introduction to RClimdex and the RClimdex tutorial (done as separate presentation from previous Workshops) and a short discussion on issues related to the comparision of station indices with indices estimated from model data. © Crown copyright 2007

What is ‘Extreme’? Wide range of space and time scales From very small scale (precip) to large scale (droughts) Definitions? High impact events Unprecedented events (in the available record) Rare events (long return periods) Exceedance of a relatively low threshold (indices, such as 10th percentile of daily temperature or 95th percentile of daily precipitation amounts) Persistence of weather conditions (droughts) Climatic extremes (e.g. extreme seasons) The most precise definition is “high impact events”, i.e. rather than defining extremes from their spatial or temporal features, it is probably better to have an “impact-based” definition. However, mostly for statistical purposes (i.e. to increase the number of events in the sample), also exceedance of low threshold are included.

Definition of 28 core extreme indices CCl/CLIVAR/JCOMM Expert Team on Climate Change Detection and Indices (ETCCDI) from ETCCDI: Definition of 28 core extreme indices Organization of regional workshop WMO-guide on extremes, 2009, targeted at NMHSs around the world http://cccma.seos.uvic.ca/ETCCDI The first part of the presentation advertises the activities of ETCCDI, the aim is to invite the participants to contribute their data to this project. These slides have been taken from some available ETCCDI presentations. Additional information about the project, extra documentation and software, including RClimdex, are available at the ETCCDI website

ETCCDI indices Internationally coordinated core set of 28 descriptive indices describe frequency, amplitude, and persistence of moderate extremes More details on the indices and the way they are estimated

Extremes Indices – temperature based ID Indicator name Definitions UNITS FD0 Frost days Annual count when TN(daily minimum)<0ºC Days SU25 Summer days Annual count when TX(daily maximum)>25ºC ID0 Ice days Annual count when TX(daily maximum)<0ºC TR20 Tropical nights Annual count when TN(daily minimum)>20ºC GSL Growing season Length Annual (1st Jan to 31st Dec in NH, 1st July to 30th June in SH) count between first span of at least 6 days with TG>5ºC and first span after July 1 (January 1 in SH) of 6 days with TG<5ºC TXx Max Tmax Monthly maximum value of daily maximum temp ºC TNx Max Tmin Monthly maximum value of daily minimum temp TXn Min Tmax Monthly minimum value of daily maximum temp TNn Min Tmin Monthly minimum value of daily minimum temp TN10p Cool nights Percentage of days when TN<10th percentile TX10p Cool days Percentage of days when TX<10th percentile TN90p Warm nights Percentage of days when TN>90th percentile TX90p Warm days Percentage of days when TX>90th percentile WSDI Warm spell duration indicator Annual count of days with at least 6 consecutive days when TX>90th percentile CSDI Cold spell duration indicator Annual count of days with at least 6 consecutive days when TN<10th percentile DTR Diurnal temperature range Monthly mean difference between TX and TN APPENDIX A: List of ETCCDMI core Climate Indices

Extremes Indices – precip based APPENDIX A: List of ETCCDMI core Climate Indices Extremes Indices – precip based ID Indicator name Definitions UNITS RX1day Max 1-day precipitation amount Monthly maximum 1-day precipitation Mm Rx5day Max 5-day precipitation amount Monthly maximum consecutive 5-day precipitation SDII Simple daily intensity index Annual total precipitation divided by the number of wet days (defined as PRCP>=1.0mm) in the year Mm/day R10 Number of heavy precipitation days Annual count of days when PRCP>=10mm Days R20 Number of very heavy precipitation days Annual count of days when PRCP>=20mm Rnn Number of days above nn mm Annual count of days when PRCP>=nn mm, nn is user defined threshold CDD Consecutive dry days Maximum number of consecutive days with RR<1mm CWD Consecutive wet days Maximum number of consecutive days with RR>=1mm R95p Very wet days Annual total PRCP when RR>95th percentile R99p Extremely wet days Annual total PRCP when RR>99th percentile mm PRCPTOT Annual total wet-day precipitation Annual total PRCP in wet days (RR>=1mm)

Alexander et al., JGR, 2006; also in IPCC, 2007 With extreme temperatures, instead, trends are larger scale for increasing temperatures. These two indices are based on the frequency of the extremes

Example: Calculating TX90p (warm days) Calculate threshold exceeded by the 10% hottest days (Tmax) in baseline period (i.e. 1961-90) On average, in the baseline period, 10% of days (36/37 days will exceed this threshold) 10% days exceed 23.2º (av. 36 days per year) 23.2º 1961 1990

Example: Calculating TX90p (warm days) Calculate the average number of times that same threshold is exceeded in a future period 58% days exceed 23.2º (av. 212 days per year) 23.2º 2070 2100 (these are synthetic data, not from real projections!)

R95PTOT- Total annual rainfall on heavy rain days Similarly, calculate the 95th percentile of wet days only (5% wettest ‘wet days’, i.e. days >1mm) in baseline These are‘ heavy rainfall days’ Calculate the average amount of rain per year that occurs in ‘heavy’ events. 12.7mm

R95PTOT- Total annual rainfall on heavy rain days Identify the ‘heavy’ rainfall days in the future Sum the rainfall that falls on those days to give average per year. 12.7mm 2070 2100

Calculating Indices with CDO To calculate some of these indices with CDO for whole model fields we can either use CDO defined extremes operators, or our own. For the percentage of warm days: cdo eca_tg90p ifile1 ifile2 ofile cdo timsum –gt ifile1 ifile2 ofile cdo mulc,100 –divc,[days] ofile ofile.percent Some of the CDO extremes operators are not always robust with PRECIS data, but we can calculate them by using other CDO operators together.

From global to local climate …. A word of warning on Validating Extremes The comparison with models data: model grid-box output as areal average of different local conditions. … from a GCM grid to the point of interest.

Individual station vs. area averages 26 stations in a 25km×25km area (black bars) and their area averages, (red bars). The area average (c.f. model grid box output) is considerably and inconsistently different to most individual stations inconsistently means that it is impossible to reconstruct the actual station time series from their relationship with the areal average time series, since there are a component due to noise (or small scale variability).

Model grid box vs. point observations Average(Extreme) ≠ Extreme(Average) Rules of thumb: Usually model output has reduced range of values and reduced variability, but it depends on the physiography of the grid box Trends should be the same if dependent on large scale phenomena (e.g. major mode of variability, climate change) included in the model and observational world Average(Extreme) /= Extreme(Average) is a result surely valid for precipitation

Acknowledgements: John Caesar (Met Office), ETCCDI Questions Acknowledgements: John Caesar (Met Office), ETCCDI © Crown copyright Met Office