In California, for example, it is well known that the forests are unhealthy and in need of more prescribed burns and other thinning efforts. However, given California’s 33 million acres of forest land, more than half of it publicly owned, even an ambitious effort like addressing the needs of a million acres a year would require decades to fix the problem.
managing the landscape
We as a society need to decide how we can restore our forests, and start a conversation about what that looks like"
“我们知道让我们的森林回到健康的状态将是未来应对火灾的最有效的方式，”Jessica Block表示，加利福尼亚大学San Diego大学Wifire实验室的业务计划副主任。“However, massive fires are destroying the ability of forests to recover."目标不是停止野火，但要理解火的作用作为管理景观的自然过程的一部分。
“We as a society need to decide how we can restore our forests, and start a conversation about what that looks like,”添加块。“我们应该将森林视为我们所在的系统，以及我们应该能够生活的系统。了解系统是目标，以便我们将来能够制定所有正确的决策。”
identify and control wildfires
西文pecially troubling is the impact of wildfires at the so-called wildland-urban interface (WUI), where growing population centers border on wildlands at risk of fire. Current fire models are not designed for these areas, so more work is needed to address these specific risks. Almost everyone agrees that the solution is to identify and control wildfires at early stages before they get out of control and turn into huge fires that impact millions of acres.
Today, postings on social media are an early warning sign but may not identify the exact location of a fire
New technologies are helping to identify nascent wildfires. One option is the addition of automatic detection capabilities to the AlertWildfire network of cameras that currently keeps watch throughout five Western states to provide early warning of wildfires. So far, human volunteers have been used to track the cameras, but automation is on the horizon. One application of machine learning is to detect a smoke flume.
A critical element is the ability to tell the difference between smoke and clouds, which humans can easily differentiate but is difficult to automate. With machine learning, computers should be able to “learn” the difference. Soon, mechanisms will exist to detect the location of a fire via multiple inputs - web cameras, social media and satellite images. Today, postings on social media are an early warning sign but may not identify the exact location of a fire.
一起工作,其他工具可以帮助pinpoint the location. Alerts to fire dispatchers must be verified as real to avoid misuse of resources.