Earthdata Forum

Dear MODIS Science Team,

I am Aakash, a researcher at Indian Institute of Space Science and Technology, working on aerosol–cloud interactions. While analyzing quasi-collocated observations from MODIS-AQUA and CALIPSO, I encountered two issues for which I seek your guidance.

Issue 1: Missing MODIS Cloud Property Retrievals

Why are cloud properties (CER, COT, LWP, CTP, CTT) often missing in MYD06 at locations where CALIPSO detects low-level single-layer clouds?
As shown in Figures 1–3, several such scenes display successful CALIPSO retrievals but missing MODIS values (fill values: −9999 or beyond the limit of valid range specified in file specification of MYD06).

Methodology:

Cloud selection: CALIPSO pixels with Mono Cloud Layer and Cloud Top Altitude < 3 km.

MODIS QA: Only valid, highest-confidence Level-2 retrievals used.

Collocation: All valid MODIS pixels within 20 km radius of CALIPSO footprint were considered.

Aggregation: CER, COT, LWP (1 km) averaged to 5 km to match geolocation.

Time matching: Daytime overpasses only, during A-Train period (2012-2015).

Sample: More than 50% points for the above described retrievals are missing (Fig 1) across 324 CALIPSO over South Asia.

I would appreciate your insights on the reasons why MODIS retrievals may be withheld despite CALIPSO identifying single-layer clouds and how to overcome this.

Issue 2: Poor CTP Agreement Between MODIS and CALIPSO

Why is the Cloud Top Pressure correlation between MODIS and CALIPSO weak?
Despite using high-quality, spatially and temporally quasi-collocating CALIPSO and MODIS data (same method as above; where all CTP_MODIS retrievals within 20km radius are averaged), the relationship is weak (see Fig. 4).

Figures (Figs 1-4) are attached in single PDF file .

Please feel free to contact me for any further clarification or if additional information.

Thank you.

Best regards,
Aakash
IIST, Thiruvananthapuram, India
Email: ak.quantun137@gmail.com

Issue 1: Missing MODIS Cloud Property Retrievals
- MYD06 and CALIPSO use different algorithms: (1) MODIS depends on visible, infrared, and TIR bands to retrieve cloud properties. (2) CALIPSO uses lidar and radar to detect cloud properties.
- MYD06 might fail to retrieve cloud properties in some cases because it is less sensitive to low-level clouds, especially in regions with poor signal contrast, such as thin clouds or clouds close to the surface.
- COT depends on the thickness and optical properties of the cloud. For Low-level clouds or thin clouds, the contrast between the cloud and the underlying surface is insufficient to provide a strong signal in TIR bands to derive accurate CER, COT, and other cloud properties.
- Similarly, for LWP and CTP: when cloud depth is shallow or the cloud is near the surface, or there’s insufficient thermal contrast, it’s difficult to estimate accurately LWP and CTP for low-level clouds.
- If CALIPSO detects low-level clouds in areas where MYD06 data is unreliable (e.g., over ocean, polar regions, or during certain seasonal conditions), it’s possible MYD06 has missing values.

Issue 2: Poor CTP Agreement Between MODIS and CALIPSO
- MODIS CTP uses a cloud-top temperature-based method and cannot directly measure the cloud’s vertical profiles. It derives CTP according to surface properties and the observed temperature of the cloud top. It is heavily affected by the surrounding atmosphere, especially if the cloud top is not a distinct temperature boundary, or if the cloud is thin. This method also depends on the clear-sky assumptions and may not be held for complex clouds. If cloud has multiple layers or is thin, MODIS might not be able to get accurate CTP, which leads to discrepancies when comparing to CALIPSO’s direct measurements.
- CALIPSO directly measures cloud vertical profiles using lidar. Its CTP is based on a high-resolution vertical cloud profile, where the cloud’s altitude is directly measured.
- MODIS is more sensitive to high-level clouds (e.g., cirrus), where there’s a clearer temperature contrast. But for low-level clouds (e.g., stratus or shallow cumulus), there can be a weak temperature signal, leading to uncertainty in the CTP retrieval.
- Difference in spatial resolution may also cause poor agreement, because MODIS may average over heterogeneous cloud fields, leading to less accurate CTP.
estimates.

Therefore, it’s better to use CALIPSO for detailed vertical structure and use MODIS for broad cloud properties.