To measure drought events, this dataset uses the Standardised Precipitation-Evapotranspiration Index (SPEI), which incorporates temperature and precipitation. We report on drought frequency, duration, severity, and intensity at three different time scales, short-term (3 months), medium-term (6 months) and long-term (12 months). These different time scales are approximately equivalent to meteorological, agricultural, and hydrological drought, respectively. We report the trends for 30 sites across Aotearoa New Zealand monitored by NIWA (National Institute for Water and Atmospheric Research) from 1972 to 2022.
Variables:
site: Site the NIWA climate stations represent.
time_scale: The number of months of drought
drought_type: The drought the SPEI values represent given at 3, 6, and 12 months (meterological, agricultural and hydrological respectively.
lat: Approx. lattitude location of NIWA climate stations to represent a site.
lon: Approx. longitude location of NIWA climate stations to represent a site.
trend_type: Duration is the number of months a drought event lasts.
Average SPEI is the annual average SPEI value. Severity is the sum of SPEI values per drought event. Intensity is severity/duration. Peak month is the lowest SPEI value recorded per drought event. Frequency is the numbers of months between each drought event.
p_value: Probability of obtaining test results at least as extreme as the result actually observed.
z: Z statistic after correcting for autocorrelated data
method: The type of trend test undertaken. Note that for methane a linear model with a quadratic term is used. For the Mann Kendall test we used a modified Mann Kendall test for autocorrelated data modifiedmk::mmkh()
n: Number of data points included in trend calculation.
note: Linear model analysis notes
s, var_s, tau: Mann-Kendall test statistics.
alternative: Alternative hypothesis
trend_likelihood: Likelihood of trend direction adapted from IPCC criteria.
period_start, period_end: The period the trend represents.
