India and the US can look into advanced methods for flood and flash flood predictions.
On the subject of flood forecasting, we can draw a number of useful parallels between India and the United States of America. Governments of each country support weather forecasting efforts that are remarkably similar, even to the numerical models employed. Each country has a weather radar network that covers much of the population, especially the urban centres. Each country’s scientific advancement on forecasting goals is both supported and hindered by democratic bureaucracy, even while citizens and their representatives plead for greater accuracy and better communication of the risks and anticipated impacts of extreme events. Each country watches its coastlines intently for the next landfalling storm event, with atmospheric rivers and hurricanes in the US that can sometimes rival the well-known southwest monsoon, Bengal cyclones, and northeast monsoon of the Indian subcontinent.
And yet this year, both countries saw devastating floods in other areas known to be inherently risky. Mountainous areas in western US and in northern India saw extreme rainfall and flooding– events brought about by the interaction of weather systems with complex terrain. Such areas are not often covered well by weather radar networks in either country, due to the physical and practical limitations of the radar systems, their locations, and their operation. A radar system situated to warn the residents of Denver, Colorado, of an approaching storm is not necessarily positioned for the observation of a widespread storm that stays in the mountains and valleys of the northern Colorado Front Range. A radar system situated in Delhi is not as useful for tracking storms over Uttarakhand, at the edge of its observation range and blocked by numerous high ridge lines in the Himalaya.
Still, weather radar and the combination of its data with rain gauges at the surface remain our most promising tools for observation and forecasting of flood events, even in complex terrain. Hydrologic forecasting with these data came about in the US during the 1990s and has continually improved with better radar systems, better methods for obtaining accurate rainfall amounts, and improved hydrologic models at the surface. These models have evolved from a lumped (watershed) basis to distributed (cell-based) methods, similar to the weather models that are also now employed to drive hydrologic forecasting. Ongoing research continues to provide better ideas about land cover and its changes, surface streams and their connections, the behaviour of water in the soil, the impact of urbanisation on runoff and downstream locations, and the particular behaviour of flood-producing storms and runoff in areas of mountainous terrain.
Where radar observations and surface precipitation gauges are lacking, hydrologic forecasting must rely on the next-best data source: weather models. No local information was available for the flood that occurred in Leh in 2010, but other sources might have proven useful with timely and appropriate application to a flood forecasting and warning effort. While driven by almost innumerable observations from around the world, including satellite observations of the entire globe, the local dynamics of a weather system may still not be well represented in models with coarse resolution. The key here is the accurate representation of both terrain and the land cover types. If the model cannot represent these well, and thus their impacts on winds and clouds and storm patterns, then it will not forecast accurately the distribution of rainfall over the area of interest. As the monsoon brings a line of storms to the edge of the Himalaya, or an atmospheric river streams from the Gulf of Mexico toward the Colorado Front Range, the accuracy of any hydrologic and flood forecast depends critically on knowing when and where that rain will fall. As we know well from tracking of hurricanes and cyclones, accurate weather forecasts provide a lead-time for flood preparation and warning that weather radar cannot. The same is true, but presents an even greater challenge, in mountainous regions.
In locations such as India and the US, operational forecasting and public warnings are a role of the federal government. While subject to the ebb and flow of research and operational funding in politicised budget cycles, activities in both countries can look to deep reserves of academic research and innovation in data collection, analysis, and model advancement. In the US, funding of data collection with taxpayer dollars has meant generally open access to data products, with a growing data infrastructure and internet-based distribution methods that provide free datasets and descriptions upon request. Freely available datasets (and their counterpart analysis in open-access publications) are a boon to academic research, where innovations are born and often returned to the public or, sometimes, incorporated as private enterprises. However, the enterprise model for forecasting stands outside of the government’s role in ensuring public health and safety. A recent study in the US accounted weather-related impacts on the national economy totalling hundreds of billions of dollars per year, with nearly a dozen economic sectors relying on government-provided weather data and services in order to take advantage of good conditions and mitigate losses. There is a growing private sector industry in weather and related services, but these still rely on the government role in essential data collection and dissemination. Where private companies contribute to the health and safety of the public, as many do (consider the very public face of The Weather Channel in the US), their added value is clear. However, it is useful to remember that they would not exist without the massive federal resources supporting new and ongoing data collection efforts (consider international satellite constellations and national radar networks) and widespread academic research programs. The beneficial value of weather forecasts far exceeds any government and private expenditures on generation and continual improvement of those forecast methods.
For further inspiration, the US and India can look to yet another area where advanced methods for flood and flash flood prediction are continually developed. A number of recent publications from Europe, especially countries in and around the Alps, demonstrate creativity in flood forecasting methods that has not seemed apparent in academic literature from the US. Similar to the process of climate modelling for long-term predictions of natural cycles and anthropogenic impacts, researchers in Switzerland have developed an ensemble method for the blending of available radar and surface rain gauge data with weather model forecasts, in a way that accounts for a number of the uncertainties in how those data and forecasts are determined, to produce probabilistic river forecasts and flood predictions. For cities subject to floods, such probabilities are highly useful in the risk-based development of response measures and the improvement of infrastructure. It remains to be seen how individuals respond to probabilities, however: some studies have suggested that a hard forecast (the river will/won’t flood today; the cyclone will/won’t strike your town) may drive immediate preparation and action, but the weather and river forecasts are rarely that specific. Storm track and river stage forecasts have errors, and the result depends on model accuracy with a measure of risk. The question remains, how much risk will you tolerate?
Fatal error: Uncaught Error:  operator not supported for strings in /home/thinkpra/public_html/archives/wp-content/themes/layerswp/core/helpers/post.php:62 Stack trace: #0 /home/thinkpra/public_html/archives/wp-content/themes/layerswp/partials/content-single.php(81): layers_post_meta(5602) #1 /home/thinkpra/public_html/archives/wp-includes/template.php(732): require('/home/thinkpra/...') #2 /home/thinkpra/public_html/archives/wp-includes/template.php(676): load_template('/home/thinkpra/...', false, Array) #3 /home/thinkpra/public_html/archives/wp-includes/general-template.php(204): locate_template(Array, true, false, Array) #4 /home/thinkpra/public_html/archives/wp-content/themes/layerswp/single.php(20): get_template_part('partials/conten...', 'single') #5 /home/thinkpra/public_html/archives/wp-includes/template-loader.php(106): include('/home/thinkpra/...') #6 /home/thinkpra/public_html/archives/wp-blog-header.php(19): require_once('/home/thinkpra/...') #7 /home/thinkpra/public_html/archives/index.php(17): require('/ho in /home/thinkpra/public_html/archives/wp-content/themes/layerswp/core/helpers/post.php on line 62