Methods
HyStat offers a wide range of statistical methods:
1. Time series and extreme value extraction
| Subject |
Method |
Screenshot |
| Time series |
Daily mean series MQ(d) aggregated to weeks MQ(w),
months MQ(m), seasons and years MQ(j) |
Interface
Graphic |
| Extreme values |
Extraction of flood series (HQ), low water series (NMxQ),
series for the duration and deficit when falling below a
threshold (MaxD, SumD, MaxV, SumV), anual and partial
extraction of series |
Interface
Graphic |
|
|
|
2. Analysis |
| Subject |
Method |
Screenshot |
| Probability |
21 different probability distributions covering
normal-, gamma-, extreme value- and other distributions,
3 parameter estimation methods (moments, L-moments und
maximum likelihood), 4 smooth tests (K-S-test,
n-omega-test, quantile correlation, chi-square test) |
Interface
Graphic |
| Trend |
Trend estimation by minimizing the error sum of
squares as well as by resistant line estimation,
Mann-Kendall-Q-test |
Interface
Graphic |
| Jump |
Jump analysis by Bernier, Pettit and the local
jump-method |
Interface
Graphic |
| Seasonality |
Parde, Grimm, seasonality index by Burn |
Interface
Graphic |
| Homogeneity |
Homogeneity check by the absolute homogeneity tests
outliers, Cox-Stuart, Kruskal-Wallis, Noether, K-S-test,
chi-square-test and cumulative sum analysis as well as
by relative homogeneity tests, double sum analysis,
Alexanderson and Buishand |
Interface
Graphic |
| 2d-probability |
2-dimensional probability analysis using normally
distributed marginal distribution |
Graphic
|
| Time series |
Autocorrelation function, partial autocorrelation
function, cross-correlation function, effective sample
size, spectral analysis, cross-spectral analysis,
regression |
Interface
Graphic |
|
|
