Page 76 - 全球气候变化及其影响Global Climate Change and Its Impacts-185×260

Global Climate Change and Its Impacts

III. Quantification and Expression of Uncertainties in Climate Change
Evidence

In the field of climate change research, uncertainties are pervasiveand profoundexist-
ing across variouscategories of evidence.within these domains, permeating multiple criti-
cal stages including data acquisition, model simulation, and the understanding of complex
climate systems. Given that climate change involves numerous complex factors - such as
atmospheric circulation, oceanic heat transport, and terrestrial ecosystem feedback - with in-
tricate interconnections and mutual influences between components, uncertainty has become
an indispensable characteristic of climate change research. Conducting precise quantitative
analysis and appropriate expression of these uncertainties holds vital significance for accu-
rately communicating climate change information, scientifically assessing climate impacts,
and formulating rational response strategies.
At the data level, error range is one of the common and effective methods for quan-
tifying uncertainty. Taking meteorological station observation data as an example, due to
instrument accuracy limitations, there is often a deviation between measured values and true
values. For instance, common mercury thermometers with minimum graduations of 0.1°C in-
herently limit measurement precision. Environmental interference also affects measurement
outcomes - the urban heat island effect can cause temperature measurements at urban weath-
er stations to be higher than the actual temperatures in surrounding natural environments.
Additionally, human operational errors, such as inaccurate readings by observers or record-
ing mistakes, may lead to deviations in measured values. Conducting repeated measurements
and calculating the standard deviation of measurement data through statistical methods can
establish a reasonable error range. For example, when measuring the annual average tem-
perature of a region, after continuous multi-year observations, if the calculated standard
deviation is ±0.3°C, the error range for the annual average temperature measurement in that
region can be expressed as the mean value ±0.3°C. This indicates a high probability that
the true annual average temperature falls within this range. For satellite data, there are also
multiple complex sources of error. Sensor measurement errors are constrained by hardware
performance and design principles - charge-coupled devices (CCD) generate inherent noise
when converting optical signals to electrical signals, affecting measurement accuracy. Radio-
metric calibration errors stem from uncertainties in converting raw radiation values measured
by satellite sensors into physical quantities (such as reflectance and radiance). Atmospheric
correction errors arise from inaccurate estimations of atmospheric components (gas composi-
tion, aerosol content, etc.), leading to errors when correcting atmospheric effects on satellite
observation data. Through comprehensive quality assessment and in-depth error analysis of
satellite data, corresponding error ranges can also be determined. When utilizing satellites to
monitor global vegetation cover changes, due to sensor resolution limitations and uncertain-
ties in vegetation type identification, vegetation coverage measurements may have an error
range of ±5%, clearly demonstrating the uncertainty level in satellite monitoring data.

• 68 •