Climate variability and relationships between top‐of‐atmosphere radiation and temperatures on Earth

2015 
The monthly global and regional variability in Earth's radiation balance is examined using correlations and regressions between atmospheric temperatures and water vapor with top-of-atmosphere outgoing longwave (OLR), absorbed shortwave (ASR), and net radiation (RT = ASR − OLR). Anomalous global mean monthly variability in the net radiation is surprisingly large, often more than ±1 W m−2, and arises mainly from clouds and transient weather systems. Relationships are strongest and positive between OLR and temperatures, especially over land for tropospheric temperatures, except in the deep tropics where high sea surface temperatures are associated with deep convection, high cold cloud tops and thus less OLR but also less ASR. Tropospheric vertically averaged temperatures (surface = 150 hPa) are thus negatively correlated globally with net radiation (−0.57), implying 2.18 ± 0.10 W m−2 extra net radiation to space for 1°C increase in temperature. Water vapor is positively correlated with tropospheric temperatures and thus also negatively correlated with net radiation; however, when the temperature dependency of water vapor is statistically removed, a significant positive feedback between water vapor and net radiation is revealed globally with 0.87 W m−2 less OLR to space per millimeter of total column water vapor. The regression coefficient between global RT and tropospheric temperature becomes −2.98 W m−2 K−1 if water vapor effects are removed, slightly less than expected from blackbody radiation (−3.2 W m−2 K−1), suggesting a positive feedback from clouds and other processes. Robust regional structures provide additional physical insights. The observational record is too short, weather noise too great, and forcing too small to make reliable estimates of climate sensitivity.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    55
    References
    57
    Citations
    NaN
    KQI
    []