Real time forecasting of the Bay of Bengal cyclonic storm “RASHMI” of October 2008 – A statistical-dynamical approach

A four-step statistical-dynamical approach is applied for real time forecasting of the Bay of Bengal cyclonic storm “RASHMI” of October 2008 which made landfall near Khepupara (Bangladesh) around 2200 UTC of 26 October 2008. The four-step approach consists of (a) Analysis of Genesis Potential Parameter (GPP), (b) Track prediction, (c) Intensity Prediction by Statistical Cyclone Intensity Prediction (SCIP) model and (d) Prediction of decaying intensity after the landfall. The results show that the analysis of Genesis Potential Parameter (GPP) at early stages of development strongly indicated that the cyclone “RASHMI” had enough potential to reach its cyclone stage. The 48 hours landfall forecast position error based on 0000 UTC on 25 October shows that the error varies from around 10 km to 95 km and landfall time error varies from 12 hours early to 23 hours delay by different numerical models (NWP). The consensus forecast (ensemble) based on these NWP models shows that landfall forecast position error is around 10 km and landfall time error is around 2 hours delay. The updated 24 hours forecast based on 0000 UTC of 26 October shows improvement in the forecast. The model predicted landfall position error varies from around 10 km to 55 km with landfall time 6 hours early to 3 hours delay. The Multiple Model Ensemble (MME) forecast shows that the landfall forecast position is close to observed landfall point and the landfall time is early by 2 hours. The JMA (Japan Meteorological Agency) and ensemble forecasts are found to be consistent both in terms of 24-hourly forecasts position, landfall point and landfall time. The 12–hourly intensity prediction up to 24 hours forecasts based on 0000 UTC on 26 October show that the model (SCIP) could pick up the intensification of the cyclone. The model forecasts till the landfall point show that there is an underestimation of intensity by 2 knots and 8 knots at 12 hour and 24 hour forecasts respectively. The 6-hourly decaying intensity forecast after the landfall shows an overestimation of 6 knots and 10 knots at 6-hour and 12-hour forecasts respectively. The approach provided useful guidance to the forecasters for real time forecasting of the cyclone.

Modelling the damage costs of invasive alien species

The rate of biological invasions is growing unprecedentedly, threatening ecological and socioeconomic systems worldwide. Quantitative understandings of invasion temporal trajectories are essential to discern current and future economic impacts of invaders, and then to inform future management strategies. Here, we examine the temporal trends of cumulative invasion costs by developing and testing a novel mathematical model with a population dynamical approach based on logistic growth. This model characterises temporal cost developments into four curve types (I–IV), each with distinct mathematical and qualitative properties, allowing for the parameterization of maximum cumulative costs, carrying capacities and growth rates. We test our model using damage cost data for eight genera (Rattus, Aedes, Canis, Oryctolagus, Sturnus, Ceratitis, Sus and Lymantria) extracted from the InvaCost database—which is the most up-to-date and comprehensive global compilation of economic cost estimates associated with invasive alien species. We find fundamental differences in the temporal dynamics of damage costs among genera, indicating they depend on invasion duration, species ecology and impacted sectors of economic activity. The fitted cost curves indicate a lack of broadscale support for saturation between invader density and impact, including for Canis, Oryctolagus and Lymantria, whereby costs continue to increase with no sign of saturation. For other taxa, predicted saturations may arise from data availability issues resulting from an underreporting of costs in many invaded regions. Overall, this population dynamical approach can produce cost trajectories for additional existing and emerging species, and can estimate the ecological parameters governing the linkage between population dynamics and cost dynamics.

Uncertainty analysis of a rainfall threshold estimate for stony debris flow based on the backward dynamical approach

A rainfall threshold is a function of some rainfall quantities that provides the conditions beyond which the probability of debris-flow occurrence is considered significant. Many uncertainties may affect the thresholds calibration and, consequently, its robustness. This study aims to assess the uncertainty in the estimate of a rainfall threshold for stony debris flow based on the backward dynamical approach, an innovative method to compute the rainfall duration and averaged intensity strictly related to a measured debris flow. The uncertainty analysis is computed by performing two Monte Carlo cascade simulations: (i) to assess the variability in the event characteristics estimate due to the uncertainty in the backward dynamical approach parameters and data and (ii) to quantify the impact of this variability on the threshold calibration. The application of this procedure to a case study highlights that the variability in the event characteristics can be both low and high. Instead, the threshold coefficients have a low dispersion showing good robustness of the threshold estimate. Moreover, the results suggest that some event features are correlated with the variability of the rainfall event duration and intensity. The proposed method is suitable to analyse the uncertainty of other threshold calibration approaches.