Figure 1
From: A novel grass hybrid to reduce flood generation in temperate regions
Time series of total precipitation, precipitation intensity and reduced plot scale runoff in Festulolium hybrid over a two year period.
Box A total precipitation (n = 2) time series for runoff events. Box B precipitation intensity (mm/hr) time series for runoff events. Boxes C and D mean plot scale runoff/rainfall (%) ((runoff (surface and 0–30 cm) for the event duration divided by the rainfall for the event duration)*100) (n = 3) from six grasses (Lp-Lolium perenne cv AberStar, Fp-Festuca pratensis cv Bf993, LpxFp- L. perenne x F. pratensis cv Prior, Lm-Lolium multiflorum,cv AberEpic, Fa-Festuca arundinacea cv Dovey and LmxFa-L. multiflorum x F. arundinacea var glaucescens cv 99/1). Statistical analysis of the runoff coefficients with variable time steps used a power function-the simplest correlation structure, where the power index is the absolute time in days between the assessments. Restricted maximum likelihood (REML) was used to determine if such a model was required. The models were compared using the difference between the residual deviances (−2*log likelihood). The difference in deviances is 84 on 1 degree of freedom (this is a chi-square statistic and is highly statistically significant). Standard error of differences of means (SED) is presented. Runoff measurements were made using high resolution (one min) data collection from laboratory and field calibrated tipping buckets from March 2007. Analysis of the time series datasets identified 33 individual rainfall events.
