Drudge Retort: The Other Side of the News

Drudge Retort

User Info

apparatchik

Subscribe to apparatchik's blog Subscribe

Menu

Special Features

Comments

You know I am really sick of people using flawed computer models to create results and then claim they are facts.

You mistake the media's generally piss-poor reporting for scientists claiming "they are facts".

This is just like these flawed global warming computer predictions that have failed

Clearly, you have no understanding of how the process works (and neither do the folks reporting on it, hence everyone's ignorance).

1. Meaningful datapoints are collected via observation that detail various climate attributes (temp, humidity, rainfall, moisture, wind, snow, ice, albedo, pressure, etc.) over a geographic area (region, continent, ocean, global). We'll call these OUTPUTS

2. Models are created to mathematically describe the OUTPUT data that has been observed, by plugging in known contributors to climate patterns that we'll call INPUTS (sun, angle toward sun of earth's axis, orbit and distance of moon, observed make-up of gases and particulates in the atmosphere, etc.) and having that model accurately and with high-fidelity re-produce nearly the same OUTPUT data points on its own.

3. Over the course of _years_, we verify the model by running it with observed INPUTS to see if its OUTPUTS happen to match the observed OUTPUTS from our satellites and weather stations. There is generally a good fit for these models.

4. Since the models have been shown to be generally accurate in hindcast and nowcast, researchers will attempt to FORECAST climate, by taking a BEST GUESS at what future inputs will look like in order to tease out future OUTPUTS. They even build similar models for guessing those INPUTS over time, that are generally fairly accurate in the hind- and nowcasts and use future numbers for best-guessing.

5. Since the original climate model worked for historical data and present data, it logically stands to reason that if we provide valid future INPUT data (INPUTS that actually reflect what the observed inputs will be at that point in the future), that we'll be able to have some predictive capability of climate data into the future. As it turns out, accurately predicting the value of those INPUTS in the future is really hard (how many factories will China build next year? volcanic eruptions? wild fires? cars on the road? cars off the road? gains in efficiency?).

There are a good number of different climate models that provide fairly accurate hind- and nowcast figures bast on observed inputs and they are in general agreement with one another. However, when you read that computer models are predicting future climate patterns, it generally means that several of these models have been used with a best-guess set of inputs to arrive at climate conditions in the future. The scientists will never claim that those conditions will in fact occur and they readily admit to an honest margin of error given the difficulty in providing future INPUT values that make any sense. The problem is that when the media reports on this, you lose out on the uncertainty that climate scientists openly admit.

That being said, however, when all of your models across all ranges of valid INPUTS are telling you, in no uncertain terms, that disruptive changes in climate are on the horizon -- the same models that have accurately provided hind- and nowcast climate data for YEARS -- you'd be dumb as a rock to not take notice.

Drudge Retort
 

Home | Breaking News | Comments | User Blogs | Stats | Back Page | RSS Feed | RSS Spec | DMCA Compliance | Privacy | Copyright 2014 World Readable