ECV Cloud Properties 28 

Cloud feedback is considered to be one of the most uncertain aspects of future climate projections and is primarily responsible for the wide range of estimates of climate sensitivity from models. The measurement of cloud properties is difficult because it is scene- and instrument-dependent. Passive remote sensing determines in general a ‘radiative’ height (in pressure coordinates). Observations from VIS/NIR/IR imagers, as well as from IR and MW sounders, have existed for more than 25 years. These are complemented by radiometers determining multi-angle reflection (also polarization) or measuring radiances in the O2 absorption band (nadir viewing and limb viewing), as well as by some active lidar and radar measurements.

The WCRP International Satellite Cloud Climatology Project (ISCCP) has developed a continuous record of infrared and visible radiances of derived cloud properties since 1983, utilizing both geostationary and polar-orbiting satellite data to resolve a three-hourly diurnal cycle. Additional datasets are available from various other instruments. To improve our understanding of clouds, the synergy of those observations is crucial. For monitoring, an overall strategy has to be developed, and newly developed datasets have to be thoroughly assessed.

Based on published knowledge of cloud radiative transfer and microphysics (precipitation formation), the minimum quantity list (in order of importance) for determining cloud effects on radiation is cloud amount (or cover) (CA), cloud top temperature/pressure (CTP/CTT), and cloud optical depth (COD), and, for determining whether clouds precipitate or not, cloud water path (CWP) or cloud effective radius (CRE). COD can be defined at a specific wavelength (e.g. 550nm) because it is possible to quantify the spectral dependence of optical depth accurately. Adding CRE for radiation would reduce error in COD but only by about 10 per cent. CWP or CRE is sufficient (statistically) to indicate precipitation onset because these two quantities vary together monotonically for non-precipitating clouds up to the onset of precipitation. It is possible to measure two of the three quantities, COD, CWP and CRE, from which the third can be inferred. However, COD can be used to estimate CWP, even without a CRE measurement, because the range of CRE variations is relatively small, i.e. no more than a factor of two. Actual CWP from satellite remote sensing is usually underestimated, even if CRE is measured, because the latter is measured nearer the cloud top and is not as sensitive to the larger particles near the cloud base. CWP can also be used to estimate cloud water content (CWC) because it varies monotonically with CWP, and cloud layer thicknesses for non-precipitating clouds do not vary much – i.e. only by about a factor of three but most of the mass is concentrated near the cloud base.

The following are thus recommended for this ECV:

Product A.6.1 Cloud amount (CA) Product A.6.2 Cloud top pressure (CTP) Product A.6.3 Cloud top temperature (CTT) Product A.6.4 Cloud optical depth (COD)

Product A.6.5 Cloud water path (liquid and ice) (CWP)

Product A.6.6 Cloud effective particle radius (liquid and ice) (CRE) Benefits

• Reduction in uncertainty in projections of future climate;

• Improvement in climate monitoring and climate model/reanalysis validation;

• Improvement in knowledge about the interaction between clouds, aerosols and atmospheric gases. Target Requirements Variable/ Parameter Horizontal Resolution Vertical Resolution Temporal

Resolution Accuracy Stability

CA 50km N/A 3h 0.01 – 0.05 0.003 – 0.03

CTP 50km N/A 3h 15hPa – 50hPa 3hPa – 15hPa

CTT 50km N/A 3h 1K – 5K 0.2K – 1K

COD 50km N/A 3h 10% 2%

CWP 50km N/A 3h 25% 5%

CRE 50km N/A 3h 5%–10% 1%–2%

Rationale: Targets for accuracy and stability have been determined by assuming a cloud feedback similar to a radiative forcing of about 0.3W/m2_{, which is roughly 20 per cent of current greenhouse gas (GHG)} forcing. Since, for radiative forcing, effective cloud amount weighted by cloud emissivity (and not cloud amount itself) is the relevant quantity, target ranges are given, with the lower value for optically thick and low clouds and the higher value for optically thin cirrus (cloud emissivity of 0.2).

Although clouds are one of the most critical factors in the climate system, a precise strategy for monitoring their properties has yet to be developed, and it is recognized that satellite measurements will unlikely meet all of these targets. For example, the ISCCP products, although less accurate, provide a very valuable heritage. As noted in the IP-10, further coordinated research is required, especially for exploring the synergy of different instruments.

Cloud properties involve a complex set of variables, and their uncertainties may depend on the scene (single or multi-layer clouds), on retrieval differences (day/night), and on the instrument sensitivity. Furthermore, cloud height (and therefore also temperature) is only determined at cloud top when using a space-borne active instrument, whereas passive remote sensing determines a ‘radiative’ height.

Currently achievable performance

Documented in the GEWEX radiation panel cloud assessment report (WCRP, 2012).

Requirements for satellite instruments and satellite datasets

• FCDRs of appropriate VIS/IR imager radiances as well as of IR and microwave radiances from at least two stable low Earth orbit satellites and from five geostationary satellites;

• Exploitation of operational meteorological satellite observations from LEO and GEO;

• Continuous full global coverage;

• Monthly statistics of cloud products, including distributions (e.g. cloud pressure in relation to cloud optical depth), in addition to averages and variability;

• Ongoing programme of research missions, using active and passive instruments to improve observation of cloud properties and to calibrate and characterize long-term products.

Calibration, validation and data archiving needs

• Validation against active ground-based and space-based observations (e.g. A-Train, EarthCare);

• Reference-type missions such as the proposed CLARREO, needed for intercalibration of VIS/IR imagers and IR sounders to achieve the accuracy and stability required for decadal time scales.

Adequacy/inadequacy of current holdings

• Current products are adequate for the evaluation of climate models as well as for monitoring large- scale spatial structure and regional variability such as El Nino Southern Oscillation (ENSO);

• Current products are not adequate for monitoring climate change because the existing observing system lacks homogeneity (e.g. orbital drift, change in channel spectral response).

Immediate action, partnerships and international coordination

• Continuation and refinement of products, including reprocessing of the existing geostationary and low Earth orbit satellite data record from the early 1980s onwards;

• Planning of further coordinated satellite missions, like the A-Train, to study the three-dimensional structure of clouds and to assess cloud products from passive remote sensing;

• Development of a strategy/method to better handle satellite-orbit drift;

• Development and implementation of a strategy of how to translate scientific improvements derived from research missions, to enhance the quality of the existing multi-decadal instrument records.

• Extension of the network of measurements at super-sites (e.g. ARM, GRUAN) suitable for validation of cloud properties;

• Sustaining of coordinated assessments of global long-term cloud products, as initiated by GEWEX;

• Continuation of the effort to produce long-term records of cloud parameters in co-ordination with the modelling and observation community (e.g. CFMIP);

• Coordination by GEWEX radiation panel and ITWG. Link to GCOS Implementation Plan

• [IP-10 Action A23] Continue the climate data record of visible and infrared radiances, for example

from the International Satellite Cloud Climatology Project, and include additional data streams as they become available; pursue reprocessing as a continuous activity taking into account lessons learnt from preceding research;

• [IP-10 Action A24] Research to improve observations of the three-dimensional spatial and temporal

distribution of cloud properties.

Other applications

• NWP model validation;

• Assessment of surface UV-B irradiance, with implications on health, biodiversity and agriculture;

• Surface insolation climatology for renewable energy applications.