Skip to content

Conjunction Data Messages

Conjunction Data Messages (CDMs) are collision risk assessments published by the 18th Space Defense Squadron. Each CDM describes a predicted close approach between two cataloged objects, including the time of closest approach (TCA), miss distance, and probability of collision (\(P_c\)).

Space-Track publishes CDMs through the CDMPublic request class, which uses the expandedspacedata controller. Unlike GP or SATCAT data, CDMs do not have a dedicated typed response struct -- use query_json() to receive results as a list of dictionaries (Python) or Vec<serde_json::Value> (Rust).

CDM Fields

The following table lists commonly used CDM fields for filtering and analysis:

Field Description
CDM_ID Unique CDM identifier
CREATED CDM creation timestamp
EMERGENCY_REPORTABLE Emergency reportable flag (Y/N)
TCA Time of Closest Approach
MIN_RNG Minimum range at TCA (km)
PC Probability of Collision
SAT_1_ID NORAD catalog ID of the first object
SAT_2_ID NORAD catalog ID of the second object
SAT_1_NAME Name of the first object
SAT_2_NAME Name of the second object
SAT1_OBJECT_TYPE Object type of the first object (e.g., PAYLOAD, DEBRIS)
SAT2_OBJECT_TYPE Object type of the second object
SAT1_RCS Radar cross-section of the first object
SAT2_RCS Radar cross-section of the second object
SAT_1_EXCL_VOL Exclusion volume of the first object
SAT_2_EXCL_VOL Exclusion volume of the second object

JSON-Only Responses

CDM queries return unstructured JSON. Use query_json() on the client to parse the response as list[dict] (Python) or Vec<serde_json::Value> (Rust). There is no typed CDMRecord struct.

Query Examples

The examples below demonstrate building CDM queries. Each query constructs a URL path; execute it with SpaceTrackClient.query_json().

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
import brahe as bh
from brahe.spacetrack import operators as op

# Query high-probability conjunction events (Pc > 1e-3)
# CDMPublic uses the expandedspacedata controller automatically
query = (
    bh.SpaceTrackQuery(bh.RequestClass.CDM_PUBLIC)
    .filter("PC", op.greater_than("1.0e-3"))
    .order_by("TCA", bh.SortOrder.DESC)
    .limit(25)
)

url_path = query.build()
print(f"High-probability CDMs:\n  {url_path}")

# Query CDMs for a specific satellite (e.g., ISS, NORAD 25544)
query = (
    bh.SpaceTrackQuery(bh.RequestClass.CDM_PUBLIC)
    .filter("SAT_1_ID", "25544")
    .order_by("TCA", bh.SortOrder.DESC)
    .limit(10)
)

url_path = query.build()
print(f"\nCDMs involving ISS:\n  {url_path}")

# Query upcoming conjunctions within the next 7 days
query = (
    bh.SpaceTrackQuery(bh.RequestClass.CDM_PUBLIC)
    .filter("TCA", op.inclusive_range(op.now(), op.now_offset(7)))
    .order_by("TCA", bh.SortOrder.ASC)
)

url_path = query.build()
print(f"\nUpcoming conjunctions (next 7 days):\n  {url_path}")

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
use brahe as bh;
use bh::spacetrack::{SpaceTrackQuery, RequestClass, SortOrder};
use bh::spacetrack::operators;

fn main() {
    // Query high-probability conjunction events (Pc > 1e-3)
    // CDMPublic uses the expandedspacedata controller automatically
    let query = SpaceTrackQuery::new(RequestClass::CDMPublic)
        .filter("PC", &operators::greater_than("1.0e-3"))
        .order_by("TCA", SortOrder::Desc)
        .limit(25);

    let url_path = query.build();
    println!("High-probability CDMs:\n  {}", url_path);

    // Query CDMs for a specific satellite (e.g., ISS, NORAD 25544)
    let query = SpaceTrackQuery::new(RequestClass::CDMPublic)
        .filter("SAT_1_ID", "25544")
        .order_by("TCA", SortOrder::Desc)
        .limit(10);

    let url_path = query.build();
    println!("\nCDMs involving ISS:\n  {}", url_path);

    // Query upcoming conjunctions within the next 7 days
    let query = SpaceTrackQuery::new(RequestClass::CDMPublic)
        .filter("TCA", &operators::inclusive_range(operators::now(), operators::now_offset(7)))
        .order_by("TCA", SortOrder::Asc);

    let url_path = query.build();
    println!("\nUpcoming conjunctions (next 7 days):\n  {}", url_path);
}
Output 
1
2
3
4
5
6
7
8
High-probability CDMs:
  /basicspacedata/query/class/cdm_public/PC/%3E1.0e-3/orderby/TCA%20desc/limit/25/format/json

CDMs involving ISS:
  /basicspacedata/query/class/cdm_public/SAT_1_ID/25544/orderby/TCA%20desc/limit/10/format/json

Upcoming conjunctions (next 7 days):
  /basicspacedata/query/class/cdm_public/TCA/now--now+7/orderby/TCA%20asc/format/json

1
2
3
4
5
6
7
8
High-probability CDMs:
  /basicspacedata/query/class/cdm_public/PC/%3E1.0e-3/orderby/TCA%20desc/limit/25/format/json

CDMs involving ISS:
  /basicspacedata/query/class/cdm_public/SAT_1_ID/25544/orderby/TCA%20desc/limit/10/format/json

Upcoming conjunctions (next 7 days):
  /basicspacedata/query/class/cdm_public/TCA/now--now+7/orderby/TCA%20asc/format/json

Working with CDM Results

After executing a CDM query with query_json(), each element in the returned list is a dictionary with string keys matching the field names above:

1
2
3
4
5
6
7
8
# Example: iterating over CDM results (after query execution)
for cdm in results:
    tca = cdm["TCA"]
    pc = float(cdm["PC"])
    sat1 = cdm["SAT_1_NAME"]
    sat2 = cdm["SAT_2_NAME"]
    miss_km = float(cdm["MIN_RNG"])
    print(f"{tca}: {sat1} vs {sat2}, Pc={pc:.2e}, miss={miss_km:.1f} km")

See Also