There are two kinds of artillery used by the United States Army. One is the howitzer (basically a big gun that you shoot in the sky to hit a distant target), and the other is the MLRS (multiple launch rocket system). (Purists will note that mortars are also indirect fire weapons which could be classified as artillery, but those are beyond the scope of this post). One MLRS system can fit on a C-130 transport plane and carries 6 rockets and has little armor protection (a HIMARS truck). The other carries twelve rockets and is built on the platform of the 35 ton Bradley Fighting Vehicle, which is too big for a C-130 transport, but two of which can fit on the larger C-17 transport plane.
The range of an MLRS rocket is about 60 kilometers, at which range it has an accuracy of 10-20 feet carrying 200 pounds of high explosives, comparable in accuracy to that of a "smart bomb" dropped by aircraft and smaller than most aircraft based bombs (most are at least 250 pounds) which can be good in close quarters and urban settings.
These numbers can be hard to relate to real life, so I pulled out a map of Denver to help illustrate this point for me. Suppose that there was an MLRS rocket system in Civic Center Park in Denver. How much territory could it reach accurately? Quite a bit, it turns out. Not only could it reach everything within the C-470/E-470 loop, it could also hit a target the size of a residential garage as far away as Black Hawk, Boulder, Erie, Brighton, Lochbuie, Denver International Airport, all of incorporated Aurora, Parker, or the Chatfield Resevoir. Existing mobile howitzers used by the U.S. Army have about half of this range, about 30 kilometers, which is roughly the distance from Civic Center Park to C-470/E-470, and are considerably less accurate (the design requirement for the existing systems was a target zone of 180 feet radius to about 900 feet depending on a number of factors).
Incidentally, this has not historically been true. Until World War II, and for many years after it, howitzers (i.e. big bullets) were much more accurate than rockets (which were inaccurate, but quite cheap). The accuracy of the modern systems is primarily due to the inclusion of expensive guidance systems and global positioning satellite technology.
The smaller model MLRS system costs about $5 million for the launcher, and about $30,000 each for a rocket.
What does this mean from a public policy/budget perspective? The issue is one of balancing dollars with what they buy. There are many different ways to provide ground troops with "fire support". They can have artillery in the field in the region where they are fighting. You can use helicopters carrying missiles. You can put missiles or bombs on unmanned aerial vehicles. You can have planes circle overhead with a howitzer sticking out the side. You can use fighter aircraft to drop bombs or fire missiles or other ordinance at ground targets. You can have ships fire guns or missiles at ground targets from the sea.
An MLRS system is one of the cheapest ways to provide fire support to ground troops. It is cheaper than a next generation howitzer. The Crusader mobile howitzer system, which was cancelled, was going to cost on the order of $24 million or more per one howitzer vehicle (although the ammunition would be cheaper per round (about $1,500 per round), which could be relevant if lots of artillery fire is anticipated). A cheap fighter aircraft (the F-35) would cost $30-$40 million each, plus smart bombs which are somewhat cheaper than an MLRS rocket, or air to ground missiles which are several times more expensive than an MLRS rocket. An F-22A costs $250 million each, plus the missiles or bombs, and has a small ordinance payload. Using a B-52 or existing fighter aircraft to drop bombs doesn't involve purchase costs, but involves considerable maintenance and ground support staff costs. Using a DD(X) destroyer to provide naval fire support (with an anticipated range of about 185 km, compared to about 24 km for the largest existing naval guns and about 37 km for the battleships recently decommissioned), like the howitzer, involves cheaper ammunition, but the delivery system will cost $3.5 billion or so each, and its usefulness declines as the conflict gets further from the coast. Ten percent of the purchase cost of a destroyer will buy 3,000+ MLRS rockets, which probably exceeds the number fired in anger in the entire Iraq War. You can't use a destroyer to provide timely fire support to troops in Ramah or Baghdad or Kabul, even if it is cutting edge technology.
Do you buy one more F-22A to drop bombs over battlefields or two more C-130s and 10 HIMARS?
Would it be more useful to provide fire support for an amphibious invasion with a DD(X), or with carrier based aircraft, or long range bomber aircraft? Would it be possible to modify some Marine landing craft to provide somewhat shorter range artillery closer to shore, instead of spending billions on a ship that can send rounds twice as far? Yes, Marines need air or sea based fire support in an amphibious invasion, and the service is inclined to see D-Day version 2.0 as its main reason for being. But, once that amphibious invasion is over, a couple of days to a couple of weeks after it is begun, a foothold where cheaper conventional ground based artillery be stationed will be secured, so fire support for amphibious assaults doesn't necessarily have to be well suited for a sustained engagement and can afford to be expensive for a brief period of time if its source (e.g. carrier based aircraft) has other useful purposes the rest of the time.
If a system like HIMARS can be airlifted to precisely where it needs to be with a field airstrip and can be secure enough to be readily resupplied with additional missiles as they are expended, early in a conflict, it may eliminate the need for more expensive fire support from the air for much of the conflict. It also gives the Army a comfort level to be working with resources it controls, instead of a Naval or Air Force resource it must request from another service to receive. On the other hand, troops securing that airstrip so that cargo planes can land need fire support from somewhere.