Wednesday 18 October 2017
Evaluation of (extreme) climate indices

Climate change studies involve investigation of daily meteorological variables, e.g., the future change of the number of frost days is assessed. The modification in the frequency of these events has more significant impact on many areas (e.g., human health, agriculture) than the monthly or seasonal mean changes. Therefore, their investigation is essential for developing targeted adaptation plans.

Quantification of these events is based on climate indices derived from daily minimum, average, maximum values or daily sums of meteorological variables (mainly temperature and precipitation). Climate indices often describe extreme events that are associated with statistically rare values at the lower or upper tails of the distribution functions (e.g., extremely hot days occur only a few times in a year in Hungary). However, some indices characterize frequent events, e.g., frost days happen through one-quarter of the year in Hungary.

The most commonly used indices inform us about the frequency of a variable crossing or its duration of being over/under a given threshold., e.g., on extremely hot day daily maximum temperature is at least 35 °C; dry periods are defined as consecutive days when daily precipitation is below 1 mm. Thanks to the fixed thresholds, these indices are easily quantifiable and also interpretable. However, it is not straightforward to use these indices for comparing areas with different climate because the frequency of a variable crossing a given threshold can vastly differ by regions. In order to partially resolve this problem, percentile-based indices were introduced. In this case fix percentage values (usually the lower or upper deciles [1]) of the distribution function are defined on a reference period to specify the thresholds, then the frequency of crossing this threshold is quantified for the future.

Some climate indices along with their definitions and units (including the ones we use most commonly) are summarized in Table 1. More indices with their definitions can be found at the WCRP [2] and ECA&D [3] websites.

Table 1
Some climate indices

Index Name Definition Unit
Crossing fix threshold
FD Frost days Tmin < 0 °C days
TN-10LT Extremely cold days Tmin < -10 °C days
TX0LT Winter days Tmax < 0 °C days
TN20GT Tropical nights Tmin > 20 °C days
SU Summer days Tmax > 25 °C days
TX30GE Hot days Tmax ≥ 30 °C days
TX35GE Extremely hot days Tmax ≥ 35 °C days
DTR Diurnal temperature range Tmax–Tmin °C
RR0.1 Days with precipitation above 0.1 mm Rday ≥ 0.1 mm days
RR1 Rainy days Rday ≥ 1 mm days
RR5 Days with precipitation above 5 mm Rday ≥ 5 mm days
RR10 Heavy precipitation days Rday ≥ 10 mm days
RR20 Extremely heavy precipitation days Rday ≥ 20 mm days
RX1 Maximum 1-day precipitation Max (Rday) mm
RX5 Maximum precipitation amount in 5 consecutive days Max (Rday i, i+1, i+2, i+3, i+4) mm
SDII Precipitation intensity Precipitation/number of rainy days mm/day
Duration based
CFD Maximum number of consecutive frost days Longest period, when Tmin < 0 °C days
HEAT First-degree heat wave
Second-degree heat wave
Third-degree heat wave
Tave ≥ 25 °C for 1 day
Tave ≥ 25 °C for 3 days or ≥ 27 °C for 1 day
Tave ≥ 27 °C for 3 days
GSL Growing season length Begins if Tave> 5 °C, ends if Tave < 5 °C (after July)  for 6 consecutive days days
LFD Length of frost-free period Begins with the last day in spring when Tmin < 0 °C and ends with the first day in autumn when Tmin < 0 °C days
CDD Maximum number of consecutive dry days Longest period, when Rday < 1 mm days
CWD Maximum number of consecutive wet days Longest period, when Rday ≥ 1 mm days
Percentile based
TX10p Ratio of cold daytimes Tmax < lower decile of Tmax (reference) %
TN90p Ratio of warm nighttimes Tmin > upper decile of Tmin (reference) %
TG10p Ratio of cold days Tave< lower decile of Tave (reference) %
TG90p Ratio of warm days Tave > upper decile of Tave (reference) %
WSDI Number/length of warm periods Tmax > upper decile of Tmax (reference) for at least 6 consecutive days no, days
CSDI Number/length of cold periods Tmin < lower decile of Tmin (reference) for at least 6 consecutive days no, days
SPI Standardized precipitation index (R-Ravg)/standard deviation on an annual/seasonal basis -
CEI Climate extremes index It examines the area affected by extremes, based on Tmin, Tmax, upper and lower deciles of R and Standardized Precipitation Index %

In the following, some examples for the changes of climate indices are shown for Hungary. Figure 1 depicts the annual changes of hot days, while Figure 2 shows the autumn precipitation intensity change - both based on the results of regional climate models applied at OMSZ. According to the investigated three simulations, hot days will remarkably increase by the end of the 21st century, by at least 20, in some areas by up to 50 days with respect to the reference period. Model results also indicate the increase of autumn precipitation intensity over Hungary.


Figure 4.1

Figure 1

Changes (in days) in the annual number of hot days by 2071–2100 compared to 1971–2000 based on the
results of three model simulations of OMSZ; significant changes are marked with dots


 Figure 4.2

Figure 2
Changes (in %) in the autumn precipitation intensity by 2071–2100 compared to 1971–2000 based on the
results of three model simulations of OMSZ; significant changes are marked with dots

[1] Deciles represent one-tenth (or 10th percentile) of sorted data in an increasing order. Accordingly, 10% of the values are smaller than the lower decile and 10% of the values are larger than the upper decile.

[2] World Climate Research Programme

[3] European Climate Assessment and Dataset

Climate indices